pythorch之目标模型检测（PennFudanPed）中出现的一系列问题及解决办法

最新推荐文章于 2023-05-23 09:15:00 发布
最新推荐文章于 2023-05-23 09:15:00 发布
阅读量1.4k 收藏 4
点赞数 2
文章标签： python pytorch 深度学习 gpu
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/ross_mystery/article/details/111410284
版权

博主根据pytorch官方中文教程跑了一下基于PennFudanPed包的目标检测，陆陆续续花费了近几天的时间，总结了其中一系列出现的问题的解决办法。
在这里不得不提一句这个官方教程对初学者来说实在是不友好。

首先放上原版代码：

# Sample code from the TorchVision 0.3 Object Detection Finetuning Tutorial
# http://pytorch.org/tutorials/intermediate/torchvision_tutorial.html

import os
import numpy as np
import torch
from PIL import Image

import torchvision
from torchvision.models.detection.faster_rcnn import FastRCNNPredictor
from torchvision.models.detection.mask_rcnn import MaskRCNNPredictor

from engine import train_one_epoch, evaluate
import utils
import transforms as T


class PennFudanDataset(object):
    def __init__(self, root, transforms):
        self.root = root
        self.transforms = transforms
        # load all image files, sorting them to
        # ensure that they are aligned
        self.imgs = list(sorted(os.listdir(os.path.join(root, "PennFudanPed/PNGImages"))))
        self.masks = list(sorted(os.listdir(os.path.join(root, "PennFudanPed/PedMasks"))))

    def __getitem__(self, idx):
        # load images ad masks
        img_path = os.path.join(self.root, "PennFudanPed/PNGImages", self.imgs[idx])
        mask_path = os.path.join(self.root, "PennFudanPed/PedMasks", self.masks[idx])
        img = Image.open(img_path).convert("RGB")
        # note that we haven't converted the mask to RGB,
        # because each color corresponds to a different instance
        # with 0 being background
        mask = Image.open(mask_path)

        mask = np.array(mask)
        # instances are encoded as different colors
        obj_ids = np.unique(mask)
        # first id is the background, so remove it
        obj_ids = obj_ids[1:]

        # split the color-encoded mask into a set
        # of binary masks
        masks = mask == obj_ids[:, None, None]

        # get bounding box coordinates for each mask
        num_objs = len(obj_ids)
        boxes = []
        for i in range(num_objs):
            pos = np.where(masks[i])
            xmin = np.min(pos[1])
            xmax = np.max(pos[1])
            ymin = np.min(pos[0])
            ymax = np.max(pos[0])
            boxes.append([xmin, ymin, xmax, ymax])

        boxes = torch.as_tensor(boxes, dtype=torch.float32)
        # there is only one class
        labels = torch.ones((num_objs,), dtype=torch.int64)
        masks = torch.as_tensor(masks, dtype=torch.uint8)

        image_id = torch.tensor([idx])
        area = (boxes[:, 3] - boxes[:, 1]) * (boxes[:, 2] - boxes[:, 0])
        # suppose all instances are not crowd
        iscrowd = torch.zeros((num_objs,), dtype=torch.int64)

        target = {}
        target["boxes"] = boxes
        target["labels"] = labels
        target["masks"] = masks
        target["image_id"] = image_id
        target["area"] = area
        target["iscrowd"] = iscrowd

        if self.transforms is not None:
            img, target = self.transforms(img, target)

        return img, target

    def __len__(self):
        return len(self.imgs)

def get_model_instance_segmentation(num_classes):
    # load an instance segmentation model pre-trained pre-trained on COCO
    model = torchvision.models.detection.maskrcnn_resnet50_fpn(pretrained=True)

    # get number of input features for the classifier
    in_features = model.roi_heads.box_predictor.cls_score.in_features
    # replace the pre-trained head with a new one
    model.roi_heads.box_predictor = FastRCNNPredictor(in_features, num_classes)

    # now get the number of input features for the mask classifier
    in_features_mask = model.roi_heads.mask_predictor.conv5_mask.in_channels
    hidden_layer = 256
    # and replace the mask predictor with a new one
    model.roi_heads.mask_predictor = MaskRCNNPredictor(in_features_mask,
                                                       hidden_layer,
                                                       num_classes)

    return model


def get_transform(train):
    transforms = []
    transforms.append(T.ToTensor())
    if train:
        transforms.append(T.RandomHorizontalFlip(0.5))
    return T.Compose(transforms)


def main():
    # train on the GPU or on the CPU, if a GPU is not available
    device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')

    # our dataset has two classes only - background and person
    num_classes = 2
    # use our dataset and defined transformations
    dataset = PennFudanDataset('PennFudanPed', get_transform(train=True))
    dataset_test = PennFudanDataset('PennFudanPed', get_transform(train=False))

    # split the dataset in train and test set
    indices = torch.randperm(len(dataset)).tolist()
    dataset = torch.utils.data.Subset(dataset, indices[:-50])
    dataset_test = torch.utils.data.Subset(dataset_test, indices[-50:])

    # define training and validation data loaders
    data_loader = torch.utils.data.DataLoader(
        dataset, batch_size=2, shuffle=True, num_workers=4,
        collate_fn=utils.collate_fn)

    data_loader_test = torch.utils.data.DataLoader(
        dataset_test, batch_size=1, shuffle=False, num_workers=4,
        collate_fn=utils.collate_fn)

    # get the model using our helper function
    model = get_model_instance_segmentation(num_classes)

    # move model to the right device
    model.to(device)

    # construct an optimizer
    params = [p for p in model.parameters() if p.requires_grad]
    optimizer = torch.optim.SGD(params, lr=0.005,
                                momentum=0.9, weight_decay=0.0005)
    # and a learning rate scheduler
    lr_scheduler = torch.optim.lr_scheduler.StepLR(optimizer,
                                                   step_size=3,
                                                   gamma=0.1)

    # let's train it for 10 epochs
    num_epochs = 10

    for epoch in range(num_epochs):
        # train for one epoch, printing every 10 iterations
        train_one_epoch(model, optimizer, data_loader, device, epoch, print_freq=10)
        # update the learning rate
        lr_scheduler.step()
        # evaluate on the test dataset
        evaluate(model, data_loader_test, device=device)

    print("That's it!")
    
if __name__ == "__main__":
    main()

直接从教程给出的链接里下载的，比较保险，我自己写的改bug改了半天没给累死。。。

在学习的过程出现的问题如下：

问题1：
In references/detection/, we have a number of helper functions to simplify training and evaluating detection models. Here, we will use references/detection/engine.py, references/detection/utils.py and references/detection/transforms.py. Just copy them to your folder and use them here.

顿时我就傻了，这哪来的references/detection/？，文档里也没有给出特殊的说明。
查阅了一下，附上链接：https://github.com/pytorch/vision/tree/master/references/detection
做实验你会发现，光这3个文件还不够，因为这3个文件还依赖其他文件，依赖的文件其实就在/references/detection目录下，我们把coco_eval.py，coco_utils.py一并拷过来，我们把这5个文件都拷到当前实验代码的文件夹下。

找不到下载路径可以尝试自建同名.py文件将代码copy进去

问题2：
运行代码后发现缺少Module cocotools，发现coco_eval.py和coco_utils.py中有import cocotools，所以我们需要安装pycocotools

windows环境下安装pycocotools用：

pip install pycocotools

是无法安装成功的。

需到https://github.com/philferriere/cocoapi下载源码，解压后切换到源码的PythonAPI目录，执行：

python setup.py build_ext --inplace

python setup.py build_ext install

安装成功。

如果还有问题可以再完成以上步骤后直接试一下

pip install pycocotools

问题3：
image和mask的路径问题，官方代码给出的路径是不完善的需要自己找到PennFudanPed的所在路径进行修改

问题4：
Ubuntu下python版COCO API的安装
使用coco数据集之前需要安装官方提供的api，github链接如下：

https://github.com/cocodataset/cocoapi

安装的时候根据python版本的不同，步骤也略有不同。

首先，切换到coco api下PythonAPI路径：

cd cocoapi/PythonAPI

python 3.x版本

python3 setup.py build_ext --inplace
python3 setup.py build_ext install

配置完成后，通过测试pycocotools能否正常导入来检查cocoapi是否安装成功

import pycocotools

一切准备就绪，这时候我们运行程序，得出结果：

Epoch: [0]  [ 0/60]  eta: 0:03:04  lr: 0.000090  loss: 2.2271 (2.2271)  loss_classifier: 0.5533 (0.5533)  loss_box_reg: 0.1392 (0.1392)  loss_mask: 1.5059 (1.5059)  loss_objectness: 0.0266 (0.0266)  loss_rpn_box_reg: 0.0020 (0.0020)  time: 3.0827  data: 1.0551  max mem: 2096
Epoch: [0]  [10/60]  eta: 0:00:38  lr: 0.000936  loss: 1.3058 (1.5063)  loss_classifier: 0.3945 (0.3925)  loss_box_reg: 0.1948 (0.2093)  loss_mask: 0.7640 (0.8690)  loss_objectness: 0.0266 (0.0258)  loss_rpn_box_reg: 0.0114 (0.0096)  time: 0.7620  data: 0.0968  max mem: 3531
Epoch: [0]  [20/60]  eta: 0:00:26  lr: 0.001783  loss: 0.8923 (1.1172)  loss_classifier: 0.2117 (0.2928)  loss_box_reg: 0.1948 (0.2062)  loss_mask: 0.3432 (0.5865)  loss_objectness: 0.0137 (0.0201)  loss_rpn_box_reg: 0.0114 (0.0116)  time: 0.5322  data: 0.0011  max mem: 3531
Epoch: [0]  [30/60]  eta: 0:00:18  lr: 0.002629  loss: 0.4799 (0.8973)  loss_classifier: 0.0861 (0.2192)  loss_box_reg: 0.1391 (0.1778)  loss_mask: 0.2489 (0.4731)  loss_objectness: 0.0074 (0.0161)  loss_rpn_box_reg: 0.0102 (0.0111)  time: 0.5268  data: 0.0012  max mem: 3531
Epoch: [0]  [40/60]  eta: 0:00:11  lr: 0.003476  loss: 0.3615 (0.7688)  loss_classifier: 0.0647 (0.1814)  loss_box_reg: 0.1135 (0.1628)  loss_mask: 0.1819 (0.3998)  loss_objectness: 0.0021 (0.0127)  loss_rpn_box_reg: 0.0102 (0.0121)  time: 0.5216  data: 0.0013  max mem: 3531
Epoch: [0]  [50/60]  eta: 0:00:05  lr: 0.004323  loss: 0.3162 (0.6775)  loss_classifier: 0.0516 (0.1548)  loss_box_reg: 0.0960 (0.1452)  loss_mask: 0.1609 (0.3543)  loss_objectness: 0.0017 (0.0107)  loss_rpn_box_reg: 0.0108 (0.0125)  time: 0.5234  data: 0.0014  max mem: 3531
Epoch: [0]  [59/60]  eta: 0:00:00  lr: 0.005000  loss: 0.2682 (0.6167)  loss_classifier: 0.0368 (0.1382)  loss_box_reg: 0.0509 (0.1305)  loss_mask: 0.1599 (0.3265)  loss_objectness: 0.0017 (0.0098)  loss_rpn_box_reg: 0.0069 (0.0118)  time: 0.5002  data: 0.0013  max mem: 3531
Epoch: [0] Total time: 0:00:33 (0.5634 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:01:03  model_time: 0.2613 (0.2613)  evaluator_time: 0.0120 (0.0120)  time: 1.2606  data: 0.9853  max mem: 3531
Test:  [49/50]  eta: 0:00:00  model_time: 0.0987 (0.1022)  evaluator_time: 0.0050 (0.0062)  time: 0.1080  data: 0.0005  max mem: 3531
Test: Total time: 0:00:06 (0.1383 s / it)
Averaged stats: model_time: 0.0987 (0.1022)  evaluator_time: 0.0050 (0.0062)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.683
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.994
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.840
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.726
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.684
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.314
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.740
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.740
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.833
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.735
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.696
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.994
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.881
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.583
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.704
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.314
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.734
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.735
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.717
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.736
Epoch: [1]  [ 0/60]  eta: 0:01:41  lr: 0.005000  loss: 0.3505 (0.3505)  loss_classifier: 0.0588 (0.0588)  loss_box_reg: 0.0663 (0.0663)  loss_mask: 0.1951 (0.1951)  loss_objectness: 0.0109 (0.0109)  loss_rpn_box_reg: 0.0194 (0.0194)  time: 1.6989  data: 1.0882  max mem: 3531
Epoch: [1]  [10/60]  eta: 0:00:31  lr: 0.005000  loss: 0.2489 (0.2650)  loss_classifier: 0.0503 (0.0481)  loss_box_reg: 0.0359 (0.0428)  loss_mask: 0.1420 (0.1534)  loss_objectness: 0.0024 (0.0036)  loss_rpn_box_reg: 0.0188 (0.0171)  time: 0.6334  data: 0.1001  max mem: 3531
Epoch: [1]  [20/60]  eta: 0:00:23  lr: 0.005000  loss: 0.2476 (0.2522)  loss_classifier: 0.0435 (0.0473)  loss_box_reg: 0.0332 (0.0372)  loss_mask: 0.1420 (0.1506)  loss_objectness: 0.0018 (0.0033)  loss_rpn_box_reg: 0.0132 (0.0138)  time: 0.5307  data: 0.0012  max mem: 3531
Epoch: [1]  [30/60]  eta: 0:00:17  lr: 0.005000  loss: 0.2003 (0.2430)  loss_classifier: 0.0359 (0.0433)  loss_box_reg: 0.0209 (0.0340)  loss_mask: 0.1446 (0.1503)  loss_objectness: 0.0010 (0.0027)  loss_rpn_box_reg: 0.0105 (0.0127)  time: 0.5315  data: 0.0011  max mem: 3531
Epoch: [1]  [40/60]  eta: 0:00:11  lr: 0.005000  loss: 0.1783 (0.2339)  loss_classifier: 0.0262 (0.0400)  loss_box_reg: 0.0184 (0.0311)  loss_mask: 0.1246 (0.1484)  loss_objectness: 0.0009 (0.0025)  loss_rpn_box_reg: 0.0083 (0.0120)  time: 0.5322  data: 0.0012  max mem: 3540
Epoch: [1]  [50/60]  eta: 0:00:05  lr: 0.005000  loss: 0.1893 (0.2272)  loss_classifier: 0.0264 (0.0377)  loss_box_reg: 0.0208 (0.0291)  loss_mask: 0.1246 (0.1469)  loss_objectness: 0.0006 (0.0021)  loss_rpn_box_reg: 0.0073 (0.0113)  time: 0.5215  data: 0.0012  max mem: 3540
Epoch: [1]  [59/60]  eta: 0:00:00  lr: 0.005000  loss: 0.1924 (0.2271)  loss_classifier: 0.0302 (0.0381)  loss_box_reg: 0.0220 (0.0289)  loss_mask: 0.1351 (0.1465)  loss_objectness: 0.0008 (0.0020)  loss_rpn_box_reg: 0.0074 (0.0117)  time: 0.5166  data: 0.0012  max mem: 3540
Epoch: [1] Total time: 0:00:32 (0.5455 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1243 (0.1243)  evaluator_time: 0.0041 (0.0041)  time: 1.1192  data: 0.9898  max mem: 3540
Test:  [49/50]  eta: 0:00:00  model_time: 0.0989 (0.1003)  evaluator_time: 0.0039 (0.0039)  time: 0.1052  data: 0.0007  max mem: 3540
Test: Total time: 0:00:06 (0.1352 s / it)
Averaged stats: model_time: 0.0989 (0.1003)  evaluator_time: 0.0039 (0.0039)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.768
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.996
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.922
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.780
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.767
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.345
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.805
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.805
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.850
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.803
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.757
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.996
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.919
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.545
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.762
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.340
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.780
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.780
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.733
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.783
Epoch: [2]  [ 0/60]  eta: 0:01:45  lr: 0.005000  loss: 0.3773 (0.3773)  loss_classifier: 0.0605 (0.0605)  loss_box_reg: 0.0670 (0.0670)  loss_mask: 0.2168 (0.2168)  loss_objectness: 0.0010 (0.0010)  loss_rpn_box_reg: 0.0319 (0.0319)  time: 1.7612  data: 1.1371  max mem: 3540
Epoch: [2]  [10/60]  eta: 0:00:31  lr: 0.005000  loss: 0.1854 (0.2000)  loss_classifier: 0.0386 (0.0326)  loss_box_reg: 0.0139 (0.0229)  loss_mask: 0.1248 (0.1327)  loss_objectness: 0.0003 (0.0011)  loss_rpn_box_reg: 0.0091 (0.0108)  time: 0.6293  data: 0.1045  max mem: 3540
Epoch: [2]  [20/60]  eta: 0:00:23  lr: 0.005000  loss: 0.1822 (0.1914)  loss_classifier: 0.0293 (0.0320)  loss_box_reg: 0.0141 (0.0206)  loss_mask: 0.1167 (0.1284)  loss_objectness: 0.0002 (0.0011)  loss_rpn_box_reg: 0.0070 (0.0092)  time: 0.5167  data: 0.0013  max mem: 3540
Epoch: [2]  [30/60]  eta: 0:00:16  lr: 0.005000  loss: 0.1795 (0.1900)  loss_classifier: 0.0282 (0.0324)  loss_box_reg: 0.0148 (0.0197)  loss_mask: 0.1183 (0.1269)  loss_objectness: 0.0005 (0.0011)  loss_rpn_box_reg: 0.0069 (0.0097)  time: 0.5144  data: 0.0014  max mem: 3540
Epoch: [2]  [40/60]  eta: 0:00:10  lr: 0.005000  loss: 0.1793 (0.1920)  loss_classifier: 0.0326 (0.0329)  loss_box_reg: 0.0181 (0.0202)  loss_mask: 0.1183 (0.1271)  loss_objectness: 0.0012 (0.0013)  loss_rpn_box_reg: 0.0086 (0.0106)  time: 0.5174  data: 0.0013  max mem: 3540
Epoch: [2]  [50/60]  eta: 0:00:05  lr: 0.005000  loss: 0.1793 (0.1924)  loss_classifier: 0.0285 (0.0327)  loss_box_reg: 0.0181 (0.0196)  loss_mask: 0.1176 (0.1278)  loss_objectness: 0.0014 (0.0014)  loss_rpn_box_reg: 0.0106 (0.0109)  time: 0.5278  data: 0.0013  max mem: 3540
Epoch: [2]  [59/60]  eta: 0:00:00  lr: 0.005000  loss: 0.1705 (0.1907)  loss_classifier: 0.0285 (0.0326)  loss_box_reg: 0.0153 (0.0190)  loss_mask: 0.1176 (0.1267)  loss_objectness: 0.0004 (0.0014)  loss_rpn_box_reg: 0.0088 (0.0110)  time: 0.5497  data: 0.0013  max mem: 3540
Epoch: [2] Total time: 0:00:33 (0.5503 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1243 (0.1243)  evaluator_time: 0.0050 (0.0050)  time: 1.1101  data: 0.9798  max mem: 3540
Test:  [49/50]  eta: 0:00:00  model_time: 0.0977 (0.0986)  evaluator_time: 0.0040 (0.0037)  time: 0.1038  data: 0.0007  max mem: 3540
Test: Total time: 0:00:06 (0.1340 s / it)
Averaged stats: model_time: 0.0977 (0.0986)  evaluator_time: 0.0040 (0.0037)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.801
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.995
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.930
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.853
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.801
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.357
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.838
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.838
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.836
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.766
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.995
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.948
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.507
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.776
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.340
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.798
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.798
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.733
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.801
Epoch: [3]  [ 0/60]  eta: 0:01:33  lr: 0.000500  loss: 0.1509 (0.1509)  loss_classifier: 0.0299 (0.0299)  loss_box_reg: 0.0090 (0.0090)  loss_mask: 0.1006 (0.1006)  loss_objectness: 0.0007 (0.0007)  loss_rpn_box_reg: 0.0107 (0.0107)  time: 1.5584  data: 1.0033  max mem: 3540
Epoch: [3]  [10/60]  eta: 0:00:32  lr: 0.000500  loss: 0.1615 (0.1816)  loss_classifier: 0.0299 (0.0295)  loss_box_reg: 0.0095 (0.0147)  loss_mask: 0.1006 (0.1250)  loss_objectness: 0.0004 (0.0007)  loss_rpn_box_reg: 0.0107 (0.0118)  time: 0.6433  data: 0.0924  max mem: 3540
Epoch: [3]  [20/60]  eta: 0:00:23  lr: 0.000500  loss: 0.1546 (0.1702)  loss_classifier: 0.0197 (0.0277)  loss_box_reg: 0.0089 (0.0132)  loss_mask: 0.1055 (0.1180)  loss_objectness: 0.0005 (0.0007)  loss_rpn_box_reg: 0.0075 (0.0107)  time: 0.5471  data: 0.0013  max mem: 3884
Epoch: [3]  [30/60]  eta: 0:00:17  lr: 0.000500  loss: 0.1429 (0.1599)  loss_classifier: 0.0197 (0.0256)  loss_box_reg: 0.0068 (0.0112)  loss_mask: 0.1033 (0.1132)  loss_objectness: 0.0003 (0.0006)  loss_rpn_box_reg: 0.0062 (0.0093)  time: 0.5347  data: 0.0013  max mem: 3884
Epoch: [3]  [40/60]  eta: 0:00:11  lr: 0.000500  loss: 0.1414 (0.1624)  loss_classifier: 0.0221 (0.0264)  loss_box_reg: 0.0068 (0.0114)  loss_mask: 0.1033 (0.1147)  loss_objectness: 0.0003 (0.0006)  loss_rpn_box_reg: 0.0062 (0.0093)  time: 0.5357  data: 0.0013  max mem: 3884
Epoch: [3]  [50/60]  eta: 0:00:05  lr: 0.000500  loss: 0.1526 (0.1635)  loss_classifier: 0.0281 (0.0269)  loss_box_reg: 0.0099 (0.0119)  loss_mask: 0.1081 (0.1150)  loss_objectness: 0.0003 (0.0006)  loss_rpn_box_reg: 0.0063 (0.0091)  time: 0.5289  data: 0.0012  max mem: 3884
Epoch: [3]  [59/60]  eta: 0:00:00  lr: 0.000500  loss: 0.1536 (0.1618)  loss_classifier: 0.0240 (0.0264)  loss_box_reg: 0.0093 (0.0116)  loss_mask: 0.1101 (0.1145)  loss_objectness: 0.0005 (0.0006)  loss_rpn_box_reg: 0.0051 (0.0087)  time: 0.5214  data: 0.0013  max mem: 3884
Epoch: [3] Total time: 0:00:33 (0.5545 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1251 (0.1251)  evaluator_time: 0.0040 (0.0040)  time: 1.1082  data: 0.9781  max mem: 3884
Test:  [49/50]  eta: 0:00:00  model_time: 0.0980 (0.0996)  evaluator_time: 0.0031 (0.0039)  time: 0.1037  data: 0.0005  max mem: 3884
Test: Total time: 0:00:06 (0.1341 s / it)
Averaged stats: model_time: 0.0980 (0.0996)  evaluator_time: 0.0031 (0.0039)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.812
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.996
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.937
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.842
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.813
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.367
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.846
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.846
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.867
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.845
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.771
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.996
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.952
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.590
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.779
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.341
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.750
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.802
Epoch: [4]  [ 0/60]  eta: 0:01:32  lr: 0.000500  loss: 0.1199 (0.1199)  loss_classifier: 0.0133 (0.0133)  loss_box_reg: 0.0034 (0.0034)  loss_mask: 0.0976 (0.0976)  loss_objectness: 0.0002 (0.0002)  loss_rpn_box_reg: 0.0053 (0.0053)  time: 1.5355  data: 1.0413  max mem: 3884
Epoch: [4]  [10/60]  eta: 0:00:30  lr: 0.000500  loss: 0.1358 (0.1646)  loss_classifier: 0.0228 (0.0267)  loss_box_reg: 0.0060 (0.0125)  loss_mask: 0.1109 (0.1175)  loss_objectness: 0.0004 (0.0010)  loss_rpn_box_reg: 0.0053 (0.0069)  time: 0.6136  data: 0.0956  max mem: 3884
Epoch: [4]  [20/60]  eta: 0:00:23  lr: 0.000500  loss: 0.1551 (0.1631)  loss_classifier: 0.0253 (0.0280)  loss_box_reg: 0.0091 (0.0117)  loss_mask: 0.1086 (0.1145)  loss_objectness: 0.0004 (0.0013)  loss_rpn_box_reg: 0.0059 (0.0076)  time: 0.5322  data: 0.0012  max mem: 3884
Epoch: [4]  [30/60]  eta: 0:00:17  lr: 0.000500  loss: 0.1608 (0.1612)  loss_classifier: 0.0254 (0.0283)  loss_box_reg: 0.0097 (0.0113)  loss_mask: 0.1030 (0.1120)  loss_objectness: 0.0006 (0.0011)  loss_rpn_box_reg: 0.0095 (0.0086)  time: 0.5534  data: 0.0013  max mem: 3884
Epoch: [4]  [40/60]  eta: 0:00:11  lr: 0.000500  loss: 0.1584 (0.1650)  loss_classifier: 0.0254 (0.0283)  loss_box_reg: 0.0098 (0.0122)  loss_mask: 0.1032 (0.1146)  loss_objectness: 0.0005 (0.0010)  loss_rpn_box_reg: 0.0095 (0.0089)  time: 0.5529  data: 0.0013  max mem: 3884
Epoch: [4]  [50/60]  eta: 0:00:05  lr: 0.000500  loss: 0.1503 (0.1610)  loss_classifier: 0.0214 (0.0267)  loss_box_reg: 0.0063 (0.0113)  loss_mask: 0.1076 (0.1140)  loss_objectness: 0.0002 (0.0009)  loss_rpn_box_reg: 0.0064 (0.0081)  time: 0.5218  data: 0.0013  max mem: 3884
Epoch: [4]  [59/60]  eta: 0:00:00  lr: 0.000500  loss: 0.1397 (0.1608)  loss_classifier: 0.0205 (0.0269)  loss_box_reg: 0.0076 (0.0114)  loss_mask: 0.1067 (0.1137)  loss_objectness: 0.0002 (0.0009)  loss_rpn_box_reg: 0.0039 (0.0078)  time: 0.5180  data: 0.0013  max mem: 3884
Epoch: [4] Total time: 0:00:33 (0.5541 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1239 (0.1239)  evaluator_time: 0.0041 (0.0041)  time: 1.1031  data: 0.9742  max mem: 3884
Test:  [49/50]  eta: 0:00:00  model_time: 0.0985 (0.0987)  evaluator_time: 0.0031 (0.0036)  time: 0.1042  data: 0.0006  max mem: 3884
Test: Total time: 0:00:06 (0.1342 s / it)
Averaged stats: model_time: 0.0985 (0.0987)  evaluator_time: 0.0031 (0.0036)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.816
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.938
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.872
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.817
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.364
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.851
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.851
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.850
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.769
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.953
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.602
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.776
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.342
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.733
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.803
Epoch: [5]  [ 0/60]  eta: 0:01:35  lr: 0.000500  loss: 0.1766 (0.1766)  loss_classifier: 0.0426 (0.0426)  loss_box_reg: 0.0088 (0.0088)  loss_mask: 0.1152 (0.1152)  loss_objectness: 0.0006 (0.0006)  loss_rpn_box_reg: 0.0094 (0.0094)  time: 1.5967  data: 1.0565  max mem: 3884
Epoch: [5]  [10/60]  eta: 0:00:32  lr: 0.000500  loss: 0.1657 (0.1729)  loss_classifier: 0.0296 (0.0295)  loss_box_reg: 0.0127 (0.0149)  loss_mask: 0.1152 (0.1188)  loss_objectness: 0.0003 (0.0010)  loss_rpn_box_reg: 0.0094 (0.0087)  time: 0.6422  data: 0.0973  max mem: 3884
Epoch: [5]  [20/60]  eta: 0:00:23  lr: 0.000500  loss: 0.1635 (0.1744)  loss_classifier: 0.0296 (0.0295)  loss_box_reg: 0.0127 (0.0138)  loss_mask: 0.1102 (0.1210)  loss_objectness: 0.0003 (0.0010)  loss_rpn_box_reg: 0.0093 (0.0091)  time: 0.5440  data: 0.0013  max mem: 3884
Epoch: [5]  [30/60]  eta: 0:00:17  lr: 0.000500  loss: 0.1538 (0.1655)  loss_classifier: 0.0217 (0.0271)  loss_box_reg: 0.0097 (0.0122)  loss_mask: 0.1079 (0.1173)  loss_objectness: 0.0002 (0.0008)  loss_rpn_box_reg: 0.0071 (0.0081)  time: 0.5286  data: 0.0012  max mem: 3884
Epoch: [5]  [40/60]  eta: 0:00:11  lr: 0.000500  loss: 0.1302 (0.1593)  loss_classifier: 0.0190 (0.0257)  loss_box_reg: 0.0061 (0.0114)  loss_mask: 0.1019 (0.1139)  loss_objectness: 0.0002 (0.0009)  loss_rpn_box_reg: 0.0049 (0.0075)  time: 0.5105  data: 0.0012  max mem: 3884
Epoch: [5]  [50/60]  eta: 0:00:05  lr: 0.000500  loss: 0.1332 (0.1623)  loss_classifier: 0.0219 (0.0258)  loss_box_reg: 0.0069 (0.0118)  loss_mask: 0.0993 (0.1159)  loss_objectness: 0.0003 (0.0010)  loss_rpn_box_reg: 0.0053 (0.0077)  time: 0.5245  data: 0.0013  max mem: 3884
Epoch: [5]  [59/60]  eta: 0:00:00  lr: 0.000500  loss: 0.1569 (0.1628)  loss_classifier: 0.0268 (0.0261)  loss_box_reg: 0.0093 (0.0118)  loss_mask: 0.1137 (0.1162)  loss_objectness: 0.0004 (0.0010)  loss_rpn_box_reg: 0.0066 (0.0077)  time: 0.5501  data: 0.0012  max mem: 3884
Epoch: [5] Total time: 0:00:33 (0.5547 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:56  model_time: 0.1247 (0.1247)  evaluator_time: 0.0041 (0.0041)  time: 1.1272  data: 0.9974  max mem: 3884
Test:  [49/50]  eta: 0:00:00  model_time: 0.0990 (0.0992)  evaluator_time: 0.0040 (0.0037)  time: 0.1051  data: 0.0007  max mem: 3884
Test: Total time: 0:00:06 (0.1353 s / it)
Averaged stats: model_time: 0.0990 (0.0992)  evaluator_time: 0.0040 (0.0037)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.810
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.928
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.852
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.809
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.360
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.845
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.845
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.867
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.844
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.766
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.951
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.593
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.774
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.337
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.798
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.798
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.733
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.802
Epoch: [6]  [ 0/60]  eta: 0:01:28  lr: 0.000050  loss: 0.1075 (0.1075)  loss_classifier: 0.0126 (0.0126)  loss_box_reg: 0.0040 (0.0040)  loss_mask: 0.0850 (0.0850)  loss_objectness: 0.0024 (0.0024)  loss_rpn_box_reg: 0.0035 (0.0035)  time: 1.4716  data: 1.0033  max mem: 3884
Epoch: [6]  [10/60]  eta: 0:00:30  lr: 0.000050  loss: 0.1384 (0.1383)  loss_classifier: 0.0151 (0.0180)  loss_box_reg: 0.0064 (0.0085)  loss_mask: 0.1121 (0.1070)  loss_objectness: 0.0001 (0.0004)  loss_rpn_box_reg: 0.0035 (0.0044)  time: 0.6086  data: 0.0920  max mem: 3884
Epoch: [6]  [20/60]  eta: 0:00:23  lr: 0.000050  loss: 0.1484 (0.1500)  loss_classifier: 0.0222 (0.0223)  loss_box_reg: 0.0078 (0.0096)  loss_mask: 0.1033 (0.1109)  loss_objectness: 0.0002 (0.0006)  loss_rpn_box_reg: 0.0063 (0.0065)  time: 0.5399  data: 0.0012  max mem: 3884
Epoch: [6]  [30/60]  eta: 0:00:16  lr: 0.000050  loss: 0.1455 (0.1480)  loss_classifier: 0.0231 (0.0217)  loss_box_reg: 0.0085 (0.0093)  loss_mask: 0.1005 (0.1100)  loss_objectness: 0.0004 (0.0006)  loss_rpn_box_reg: 0.0063 (0.0064)  time: 0.5344  data: 0.0014  max mem: 3884
Epoch: [6]  [40/60]  eta: 0:00:11  lr: 0.000050  loss: 0.1432 (0.1506)  loss_classifier: 0.0201 (0.0225)  loss_box_reg: 0.0085 (0.0095)  loss_mask: 0.1027 (0.1114)  loss_objectness: 0.0004 (0.0006)  loss_rpn_box_reg: 0.0062 (0.0065)  time: 0.5402  data: 0.0014  max mem: 3956
Epoch: [6]  [50/60]  eta: 0:00:05  lr: 0.000050  loss: 0.1513 (0.1554)  loss_classifier: 0.0267 (0.0238)  loss_box_reg: 0.0096 (0.0104)  loss_mask: 0.1087 (0.1130)  loss_objectness: 0.0003 (0.0008)  loss_rpn_box_reg: 0.0071 (0.0074)  time: 0.5671  data: 0.0014  max mem: 3956
Epoch: [6]  [59/60]  eta: 0:00:00  lr: 0.000050  loss: 0.1494 (0.1569)  loss_classifier: 0.0267 (0.0242)  loss_box_reg: 0.0088 (0.0105)  loss_mask: 0.1087 (0.1138)  loss_objectness: 0.0003 (0.0010)  loss_rpn_box_reg: 0.0096 (0.0074)  time: 0.5403  data: 0.0014  max mem: 3956
Epoch: [6] Total time: 0:00:33 (0.5597 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1264 (0.1264)  evaluator_time: 0.0035 (0.0035)  time: 1.1199  data: 0.9890  max mem: 3956
Test:  [49/50]  eta: 0:00:00  model_time: 0.0981 (0.0999)  evaluator_time: 0.0035 (0.0036)  time: 0.1040  data: 0.0005  max mem: 3956
Test: Total time: 0:00:06 (0.1344 s / it)
Averaged stats: model_time: 0.0981 (0.0999)  evaluator_time: 0.0035 (0.0036)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.818
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.937
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.872
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.819
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.362
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.850
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.771
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.953
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.596
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.778
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.340
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.801
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.801
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.733
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.804
Epoch: [7]  [ 0/60]  eta: 0:01:39  lr: 0.000050  loss: 0.1236 (0.1236)  loss_classifier: 0.0174 (0.0174)  loss_box_reg: 0.0044 (0.0044)  loss_mask: 0.0943 (0.0943)  loss_objectness: 0.0001 (0.0001)  loss_rpn_box_reg: 0.0073 (0.0073)  time: 1.6640  data: 1.0842  max mem: 3956
Epoch: [7]  [10/60]  eta: 0:00:32  lr: 0.000050  loss: 0.1850 (0.1678)  loss_classifier: 0.0215 (0.0257)  loss_box_reg: 0.0152 (0.0145)  loss_mask: 0.1130 (0.1185)  loss_objectness: 0.0002 (0.0004)  loss_rpn_box_reg: 0.0090 (0.0088)  time: 0.6590  data: 0.0998  max mem: 3956
Epoch: [7]  [20/60]  eta: 0:00:24  lr: 0.000050  loss: 0.1546 (0.1658)  loss_classifier: 0.0260 (0.0263)  loss_box_reg: 0.0106 (0.0126)  loss_mask: 0.1159 (0.1177)  loss_objectness: 0.0003 (0.0005)  loss_rpn_box_reg: 0.0078 (0.0087)  time: 0.5470  data: 0.0014  max mem: 3956
Epoch: [7]  [30/60]  eta: 0:00:17  lr: 0.000050  loss: 0.1490 (0.1660)  loss_classifier: 0.0248 (0.0259)  loss_box_reg: 0.0082 (0.0123)  loss_mask: 0.1071 (0.1189)  loss_objectness: 0.0003 (0.0005)  loss_rpn_box_reg: 0.0053 (0.0084)  time: 0.5348  data: 0.0013  max mem: 3956
Epoch: [7]  [40/60]  eta: 0:00:11  lr: 0.000050  loss: 0.1490 (0.1636)  loss_classifier: 0.0265 (0.0267)  loss_box_reg: 0.0080 (0.0118)  loss_mask: 0.1039 (0.1162)  loss_objectness: 0.0003 (0.0006)  loss_rpn_box_reg: 0.0065 (0.0083)  time: 0.5369  data: 0.0013  max mem: 3956
Epoch: [7]  [50/60]  eta: 0:00:05  lr: 0.000050  loss: 0.1513 (0.1606)  loss_classifier: 0.0245 (0.0260)  loss_box_reg: 0.0077 (0.0110)  loss_mask: 0.1030 (0.1150)  loss_objectness: 0.0002 (0.0006)  loss_rpn_box_reg: 0.0062 (0.0080)  time: 0.5265  data: 0.0012  max mem: 3956
Epoch: [7]  [59/60]  eta: 0:00:00  lr: 0.000050  loss: 0.1368 (0.1580)  loss_classifier: 0.0206 (0.0254)  loss_box_reg: 0.0061 (0.0107)  loss_mask: 0.1036 (0.1136)  loss_objectness: 0.0002 (0.0007)  loss_rpn_box_reg: 0.0052 (0.0076)  time: 0.5242  data: 0.0010  max mem: 3956
Epoch: [7] Total time: 0:00:33 (0.5570 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:54  model_time: 0.1241 (0.1241)  evaluator_time: 0.0040 (0.0040)  time: 1.0991  data: 0.9689  max mem: 3956
Test:  [49/50]  eta: 0:00:00  model_time: 0.0984 (0.0986)  evaluator_time: 0.0031 (0.0036)  time: 0.1037  data: 0.0004  max mem: 3956
Test: Total time: 0:00:06 (0.1338 s / it)
Averaged stats: model_time: 0.0984 (0.0986)  evaluator_time: 0.0031 (0.0036)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.816
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.937
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.872
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.817
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.365
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.849
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.849
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.847
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.771
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.953
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.615
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.776
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.339
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.800
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.800
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.750
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.803
Epoch: [8]  [ 0/60]  eta: 0:01:28  lr: 0.000050  loss: 0.1436 (0.1436)  loss_classifier: 0.0223 (0.0223)  loss_box_reg: 0.0107 (0.0107)  loss_mask: 0.1064 (0.1064)  loss_objectness: 0.0004 (0.0004)  loss_rpn_box_reg: 0.0038 (0.0038)  time: 1.4688  data: 1.0200  max mem: 3956
Epoch: [8]  [10/60]  eta: 0:00:31  lr: 0.000050  loss: 0.1544 (0.1680)  loss_classifier: 0.0250 (0.0291)  loss_box_reg: 0.0106 (0.0131)  loss_mask: 0.1089 (0.1142)  loss_objectness: 0.0003 (0.0021)  loss_rpn_box_reg: 0.0087 (0.0095)  time: 0.6283  data: 0.0938  max mem: 3956
Epoch: [8]  [20/60]  eta: 0:00:23  lr: 0.000050  loss: 0.1445 (0.1554)  loss_classifier: 0.0245 (0.0254)  loss_box_reg: 0.0078 (0.0105)  loss_mask: 0.1049 (0.1102)  loss_objectness: 0.0002 (0.0013)  loss_rpn_box_reg: 0.0071 (0.0081)  time: 0.5385  data: 0.0013  max mem: 3956
Epoch: [8]  [30/60]  eta: 0:00:16  lr: 0.000050  loss: 0.1368 (0.1520)  loss_classifier: 0.0178 (0.0240)  loss_box_reg: 0.0072 (0.0093)  loss_mask: 0.1034 (0.1102)  loss_objectness: 0.0001 (0.0010)  loss_rpn_box_reg: 0.0045 (0.0075)  time: 0.5246  data: 0.0014  max mem: 3956
Epoch: [8]  [40/60]  eta: 0:00:11  lr: 0.000050  loss: 0.1605 (0.1587)  loss_classifier: 0.0217 (0.0238)  loss_box_reg: 0.0073 (0.0106)  loss_mask: 0.1098 (0.1156)  loss_objectness: 0.0002 (0.0009)  loss_rpn_box_reg: 0.0066 (0.0078)  time: 0.5281  data: 0.0012  max mem: 3956
Epoch: [8]  [50/60]  eta: 0:00:05  lr: 0.000050  loss: 0.1545 (0.1568)  loss_classifier: 0.0233 (0.0244)  loss_box_reg: 0.0104 (0.0107)  loss_mask: 0.1057 (0.1133)  loss_objectness: 0.0003 (0.0008)  loss_rpn_box_reg: 0.0081 (0.0076)  time: 0.5401  data: 0.0013  max mem: 3956
Epoch: [8]  [59/60]  eta: 0:00:00  lr: 0.000050  loss: 0.1442 (0.1569)  loss_classifier: 0.0220 (0.0243)  loss_box_reg: 0.0069 (0.0105)  loss_mask: 0.1057 (0.1139)  loss_objectness: 0.0002 (0.0007)  loss_rpn_box_reg: 0.0056 (0.0075)  time: 0.5459  data: 0.0013  max mem: 3956
Epoch: [8] Total time: 0:00:33 (0.5542 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1243 (0.1243)  evaluator_time: 0.0040 (0.0040)  time: 1.1066  data: 0.9773  max mem: 3956
Test:  [49/50]  eta: 0:00:00  model_time: 0.0991 (0.0992)  evaluator_time: 0.0030 (0.0035)  time: 0.1045  data: 0.0005  max mem: 3956
Test: Total time: 0:00:06 (0.1338 s / it)
Averaged stats: model_time: 0.0991 (0.0992)  evaluator_time: 0.0030 (0.0035)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.819
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.937
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.872
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.820
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.367
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.850
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.769
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.953
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.615
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.775
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.337
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.750
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.802
Epoch: [9]  [ 0/60]  eta: 0:01:36  lr: 0.000005  loss: 0.1444 (0.1444)  loss_classifier: 0.0269 (0.0269)  loss_box_reg: 0.0093 (0.0093)  loss_mask: 0.1018 (0.1018)  loss_objectness: 0.0001 (0.0001)  loss_rpn_box_reg: 0.0062 (0.0062)  time: 1.6025  data: 1.0594  max mem: 3956
Epoch: [9]  [10/60]  eta: 0:00:32  lr: 0.000005  loss: 0.1654 (0.1667)  loss_classifier: 0.0269 (0.0277)  loss_box_reg: 0.0129 (0.0115)  loss_mask: 0.1055 (0.1176)  loss_objectness: 0.0005 (0.0011)  loss_rpn_box_reg: 0.0062 (0.0088)  time: 0.6510  data: 0.0972  max mem: 3956
Epoch: [9]  [20/60]  eta: 0:00:24  lr: 0.000005  loss: 0.1354 (0.1516)  loss_classifier: 0.0241 (0.0253)  loss_box_reg: 0.0069 (0.0090)  loss_mask: 0.0978 (0.1088)  loss_objectness: 0.0002 (0.0009)  loss_rpn_box_reg: 0.0061 (0.0075)  time: 0.5623  data: 0.0011  max mem: 3956
Epoch: [9]  [30/60]  eta: 0:00:17  lr: 0.000005  loss: 0.1376 (0.1553)  loss_classifier: 0.0241 (0.0260)  loss_box_reg: 0.0077 (0.0104)  loss_mask: 0.0973 (0.1102)  loss_objectness: 0.0002 (0.0008)  loss_rpn_box_reg: 0.0062 (0.0079)  time: 0.5578  data: 0.0014  max mem: 3956
Epoch: [9]  [40/60]  eta: 0:00:11  lr: 0.000005  loss: 0.1469 (0.1578)  loss_classifier: 0.0226 (0.0253)  loss_box_reg: 0.0085 (0.0110)  loss_mask: 0.1082 (0.1129)  loss_objectness: 0.0002 (0.0008)  loss_rpn_box_reg: 0.0059 (0.0079)  time: 0.5510  data: 0.0015  max mem: 3956
Epoch: [9]  [50/60]  eta: 0:00:05  lr: 0.000005  loss: 0.1439 (0.1562)  loss_classifier: 0.0196 (0.0249)  loss_box_reg: 0.0069 (0.0107)  loss_mask: 0.1067 (0.1119)  loss_objectness: 0.0004 (0.0009)  loss_rpn_box_reg: 0.0055 (0.0077)  time: 0.5343  data: 0.0013  max mem: 3956
Epoch: [9]  [59/60]  eta: 0:00:00  lr: 0.000005  loss: 0.1454 (0.1579)  loss_classifier: 0.0194 (0.0248)  loss_box_reg: 0.0071 (0.0107)  loss_mask: 0.1067 (0.1141)  loss_objectness: 0.0004 (0.0009)  loss_rpn_box_reg: 0.0051 (0.0075)  time: 0.5102  data: 0.0012  max mem: 3956
Epoch: [9] Total time: 0:00:33 (0.5621 s / it)
creating index...
index created!
Test:  [ 0/50]  eta: 0:00:55  model_time: 0.1242 (0.1242)  evaluator_time: 0.0051 (0.0051)  time: 1.1091  data: 0.9788  max mem: 3956
Test:  [49/50]  eta: 0:00:00  model_time: 0.0997 (0.0992)  evaluator_time: 0.0030 (0.0036)  time: 0.1053  data: 0.0006  max mem: 3956
Test: Total time: 0:00:06 (0.1343 s / it)
Averaged stats: model_time: 0.0997 (0.0992)  evaluator_time: 0.0030 (0.0036)
Accumulating evaluation results...
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
IoU metric: bbox
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.819
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.937
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.872
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.820
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.367
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.852
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.883
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.850
IoU metric: segm
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.769
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.997
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.953
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.615
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.775
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.337
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.799
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.750
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.802
That's it!

Process finished with exit code 0

完成！

参考：
https://blog.csdn.net/it_boy__/article/details/106376437
https://blog.csdn.net/u013685264/article/details/100331064

ross_mystery
关注
  • 2
    点赞
  • 踩
  • 4
    收藏
    觉得还不错? 一键收藏
  • 2
    评论
目标检测算法训练数据准备——Penn-Fudan数据集预处理实例说明（附代码）
m0_49963403的博客
01-30 1780
本文以Penn-Fudan数据集预处理为例，说明用于目标检测算法训练的数据集的预处理方法及过程。
PennFudanPed行人检测数据集
11-18
PennFudanPed行人检测数据集，用于行人检测，欢迎深度学习研究者下载，欢迎交流学习，有问题请留言，
PennFudanPed数据集转yolo格式
wukong168的博客
01-26 1676
pennFudanPed数据集只有不到200张图片，标注的行人为非遮挡行人，遮挡和骑车的行人都没有标注 标注框做的比较标注和准确，下面是该数据集直接转成yolo格式的代码，新建train文件夹，然后使用下面代码存为py，然后运行 #!/usr/bin/env python # -*- coding: utf-8 -*- import os import re from pathlib import Path from PIL import Image import csv import shutil i
PennFudanPed数据集解析截取
理工男Happy大叔的博客
02-28 2978
数据集的标注文件是文本格式 # Compatible with PASCAL Annotation Version 1.00 Image filename : "PennFudanPed/PNGImages/FudanPed00001.png" Image size (X x Y x C) : 559 x 536 x 3 Database : "The Penn-Fudan-Pedestrian ...
【纯小白】动手实现MASK RCNN 实例分割（带全部源码）
m0_51325463的博客
10-25 7003
本文档依据Pytorch官方，CSDN，哔哩哔哩，Github多方总结改变实现，如有侵权，联系速删！！此文档旨在帮助小白使用pytorch一步步实现Mask Rcnn 网络。后续将会推出如何使用此网络训练自己的数据集。如在搭建过程中出现种种错误，欢迎骚扰我，我会尽力解决。
RCF-pytorch:pytorch CVPR2017中边缘检测模型的更丰富的卷积功能
04-29
丰富的边缘检测卷积功能多亏帮助由XuanyiLi创建，如果您在使用时遇到任何问题，请联系： 。 我的pytorch模型的最佳结果现在是0.808 ODS F分数。我的模型结果以下是侧面输出和预测示例引文如果您发现我们的工作对您的...
Pythorch中torch.nn.LSTM()参数详解
04-27
Pythorch中torch.nn.LSTM()参数详解；Pythorch中torch.nn.LSTM()参数详解；Pythorch中torch.nn.LSTM()参数详解；Pythorch中torch.nn.LSTM()参数详解；Pythorch中torch.nn.LSTM()参数详解；Pythorch中torch.nn.LSTM()...
pythorch mnist手写数字 CNN MLP LSTM 识别
03-27
在这个项目中，CNN可能被用来构建模型的第一部分，用于捕获图像的局部特征和空间关系。 接着，MLP（多层感知机）是一种前馈神经网络，通常用于非结构化数据的分类。在MNIST识别任务中，MLP可以处理经过展平的图像...
Colorization:彩色图像着色的pythorch实现。 在ECCV中，2016年
05-09
在ECCV中，2016年。 版本信息：pytorch 0.3.0，python 3.5 安装 git clone 培训和验证 从头开始培训： python main.py 或从指定时间点恢复训练： python main.py -R *** 以上两个命令默认都支持交叉验证，可以...
torch学习 (三十八)：使用TorchVision微调物体检测模型 (Penn-Fudan数据库)
因吉的博客
12-24 3299
本文对Penn-Fudan数据库的行人检测和分割的预训练Mask R-CNN模型进行微调，其包含170张图片，且有345个行人实例。
全卷积网络（Penn-Fudan Database数据集）
07-27
全卷积网络（Penn-Fudan Database数据集）
FUDAN-CH05 结构化分析与设计-
12-17
FUDAN-CH05 结构化分析与设计
python目标识别及实例分割代码（可以套用自己的数据）
阿利同学的博客
10-20 2350
#图像二分类#实例分割# 目标检测+实例分割
torchvision对象检测微调教程
yanglamei1962的博客
05-23 167
用于训练对象检测，实例细分和人员关键点检测的参考脚本可轻松支持添加新的自定义数据集。数据集应继承自标准类，并实现__len__和。我们唯一需要的特异性是数据集图像：大小为(H, W)的 PIL 图像目标：包含以下字段的字典格式的N边界框的坐标，范围从0至W，从0至H：每个边界框的标签。0始终代表背景类。：图像标识符。它在数据集中的所有图像之间应该是唯一的，并在评估过程中使用：边界框的区域。在使用 COCO 度量进行评估时，可使用此值来区分小盒子，中盒子和大盒子之间的度量得分。
PyTorch学习笔记(十二) ---- PyTorch目标检测
不忘初心~
08-22 5992
PyTorch学习笔记(十二) ---- PyTorch目标检测
手把手教你训练自己的Mask R-CNN图像实例分割模型（PyTorch官方教程）
热门推荐
一个菜鸟的奋斗
09-06 6万+
近来在学习图像分割的相关算法，准备试试看Mask R-CNN的效果。 关于Mask R-CNN的详细理论说明，可以参见原作论文https://arxiv.org/abs/1703.06870，网上也有大量解读的文章。本篇博客主要是参考了PyTorch官方给出的训练教程，将如何在自己的数据集上训练Mask R-CNN模型的过程记录下来，希望能为感兴趣的读者提供一些帮助。 PyTorch官方教程（...
python处理优化自己的目标检测数据集
杰斯的盐
12-05 1513
最近跑yolo的时候得到一批Penn-Fudan的行人数据集，想着处理优化后变成自己的数据放到yolo里面，会对模型的训练质量有提高。 这份数据集是开源的，需要的同学可以直接在网上搜索到。   希望能提供一些思路，如果觉得有用欢迎赞我表扬我~   首先看看拿到的数据集是什么样子的。 图片为png彩图，无固定尺寸：   标签label是txt格式，里面包含的信息如下： # C...
Object Detection finetuning tutorial + Pytorch代码实战练兵
织席贩履之辈
05-03 1187
本篇文章的学习资料来源于Pytorch官网教程内容，旨在加深对Pytorch的整体使用把握和代码细节理解，如有误读之处，请提出来，笔者愿向大家学习。 声明：英文说明我就不翻译了，直接搬过来，因为很容易看懂。 正文： For this tutorial, we will be finetuning a pre-trained Mask R-CNN model in the Penn-Fudan Da...
Pythorch如何进行模型测试及性能分析
最新发布
06-12
PyTorch中进行模型测试及性能分析通常有以下几个步骤： 1. 加载模型 使用`torch.load()`方法加载已经训练好的模型，可以使用GPU或CPU来进行模型推理。 2. 准备测试数据 准备要进行测试的数据，可以使用`torch....
评论 2
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值