2021水下目标检测记录

当前精度

0.55+

预处理

转为coco格式，同时4:1划分train val

root = '/home/yons/data/underwater2021'
import glob
xml_paths = glob.glob(root + '/' + 'xmls/*.xml')
import random
from nncv.transvert import xml2coco

random.Random(42).shuffle(xml_paths)
total_number = len(xml_paths)
index = list(range(total_number))


json_path = root + '/' + 'instances_trainval2017.json'
xml2coco(xml_paths, json_path, class_names, ignore=True)
    
val_index = index[:int(0.2*len(index))]
train_index = list(set(index) - set(val_index))
json_path = root + '/' + 'instances_train2017.json'
xml2coco([xml_paths[i] for i in train_index], json_path, class_names, ignore=True)
json_path = root + '/' + 'instances_val2017.json'
xml2coco([xml_paths[i] for i in val_index], json_path, class_names, ignore=True)

训练

跑个cas r50吧

num_classes = 4
model = dict(
    type='CascadeRCNN',
    pretrained=None,
    #  pretrained='pretrain/resnet50-19c8e357.pth',
    backbone=dict(
        type='ResNet',
        depth=50,
        num_stages=4,
        out_indices=(0, 1, 2, 3),
        frozen_stages=1,
        norm_cfg=dict(type='BN', requires_grad=True),
        norm_eval=True,
        style='pytorch',
        dcn=dict(type='DCN', deform_groups=1, fallback_on_stride=False),
        stage_with_dcn=(False, True, True, True)),
    neck=dict(
        type='FPN',
        in_channels=[256, 512, 1024, 2048],
        out_channels=256,
        num_outs=5),
    rpn_head=dict(
        type='RPNHead',
        in_channels=256,
        feat_channels=256,
        anchor_generator=dict(
            type='AnchorGenerator',
            scales=[8],
            ratios=[0.5, 1.0, 2.0],
            strides=[4, 8, 16, 32, 64]),
        bbox_coder=dict(
            type='DeltaXYWHBBoxCoder',
            target_means=[.0, .0, .0, .0],
            target_stds=[1.0, 1.0, 1.0, 1.0]),
        loss_cls=dict(
            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)),
    roi_head=dict(
        type='CascadeRoIHead',
        num_stages=3,
        stage_loss_weights=[1, 0.5, 0.25],
        bbox_roi_extractor=dict(
            type='SingleRoIExtractor',
            roi_layer=dict(type='RoIAlign', out_size=7, sample_num=0),
            out_channels=256,
            featmap_strides=[4, 8, 16, 32]),
        bbox_head=[
            dict(
                type='Shared2FCBBoxHead',
                in_channels=256,
                fc_out_channels=1024,
                roi_feat_size=7,
                num_classes=num_classes,
                bbox_coder=dict(
                    type='DeltaXYWHBBoxCoder',
                    target_means=[0., 0., 0., 0.],
                    target_stds=[0.1, 0.1, 0.2, 0.2]),
                reg_class_agnostic=True,
                loss_cls=dict(
                    type='CrossEntropyLoss',
                    use_sigmoid=False,
                    loss_weight=1.0),
                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
                               loss_weight=1.0)),
            dict(
                type='Shared2FCBBoxHead',
                in_channels=256,
                fc_out_channels=1024,
                roi_feat_size=7,
                num_classes=num_classes,
                bbox_coder=dict(
                    type='DeltaXYWHBBoxCoder',
                    target_means=[0., 0., 0., 0.],
                    target_stds=[0.05, 0.05, 0.1, 0.1]),
                reg_class_agnostic=True,
                loss_cls=dict(
                    type='CrossEntropyLoss',
                    use_sigmoid=False,
                    loss_weight=1.0),
                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
                               loss_weight=1.0)),
            dict(
                type='Shared2FCBBoxHead',
                in_channels=256,
                fc_out_channels=1024,
                roi_feat_size=7,
                num_classes=num_classes,
                bbox_coder=dict(
                    type='DeltaXYWHBBoxCoder',
                    target_means=[0., 0., 0., 0.],
                    target_stds=[0.033, 0.033, 0.067, 0.067]),
                reg_class_agnostic=True,
                loss_cls=dict(
                    type='CrossEntropyLoss',
                    use_sigmoid=False,
                    loss_weight=1.0),
                loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0))
        ]))
# model training and testing settings
train_cfg = dict(
    rpn=dict(
        assigner=dict(
            type='MaxIoUAssigner',
            pos_iou_thr=0.7,
            neg_iou_thr=0.3,
            min_pos_iou=0.3,
            match_low_quality=True,
            ignore_iof_thr=-1),
        sampler=dict(
            type='RandomSampler',
            num=256,
            pos_fraction=0.5,
            neg_pos_ub=-1,
            add_gt_as_proposals=False),
        allowed_border=0,
        pos_weight=-1,
        debug=False),
    rpn_proposal=dict(
        nms_across_levels=False,
        nms_pre=2000,
        nms_post=2000,
        max_num=2000,
        nms_thr=0.7,
        min_bbox_size=0),
    rcnn=[
        dict(
            assigner=dict(
                type='MaxIoUAssigner',
                pos_iou_thr=0.5,
                neg_iou_thr=0.5,
                min_pos_iou=0.5,
                match_low_quality=False,
                ignore_iof_thr=-1),
            sampler=dict(
                type='RandomSampler',
                num=512,
                pos_fraction=0.25,
                neg_pos_ub=-1,
                add_gt_as_proposals=True),
            pos_weight=-1,
            debug=False),
        dict(
            assigner=dict(
                type='MaxIoUAssigner',
                pos_iou_thr=0.6,
                neg_iou_thr=0.6,
                min_pos_iou=0.6,
                match_low_quality=False,
                ignore_iof_thr=-1),
            sampler=dict(
                type='RandomSampler',
                num=512,
                pos_fraction=0.25,
                neg_pos_ub=-1,
                add_gt_as_proposals=True),
            pos_weight=-1,
            debug=False),
        dict(
            assigner=dict(
                type='MaxIoUAssigner',
                pos_iou_thr=0.7,
                neg_iou_thr=0.7,
                min_pos_iou=0.7,
                match_low_quality=False,
                ignore_iof_thr=-1),
            sampler=dict(
                type='RandomSampler',
                num=512,
                pos_fraction=0.25,
                neg_pos_ub=-1,
                add_gt_as_proposals=True),
            pos_weight=-1,
            debug=False)
    ])
test_cfg = dict(
    rpn=dict(
        nms_across_levels=False,
        nms_pre=1000,
        nms_post=1000,
        max_num=1000,
        nms_thr=0.7,
        min_bbox_size=0),
    rcnn=dict(
        score_thr=0.001,
        nms=dict(type='soft_nms', iou_threshold=0.5, min_score=0.0001),
        max_per_img=200))

dataset_type = 'CocoDataset'
data_root = 'data/underwater2021/'
classes = ['holothurian', 'echinus', 'scallop', 'starfish']
img_norm_cfg = dict(
    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)

train_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(type='LoadAnnotations', with_bbox=True),
    dict(type='Resize', img_scale=[(9999, 640), (9999, 1280)], multiscale_mode='range', keep_ratio=True),
    dict(type='RandomFlip', flip_ratio=0.5),
    dict(type='Normalize', **img_norm_cfg),
    dict(type='Pad', size_divisor=32),
    dict(type='DefaultFormatBundle'),
    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']),
]
test_pipeline = [
    dict(type='LoadImageFromFile'),
    dict(
        type='MultiScaleFlipAug',
        img_scale=(9999, 960),
        flip=False,
        transforms=[
            dict(type='Resize', keep_ratio=True),
            dict(type='RandomFlip'),
            dict(type='Normalize', **img_norm_cfg),
            dict(type='Pad', size_divisor=32),
            dict(type='ImageToTensor', keys=['img']),
            dict(type='Collect', keys=['img']),
        ])
]

data = dict(
    samples_per_gpu=2,
    workers_per_gpu=2,
    train=dict(
        classes=classes,
        type=dataset_type,
        ann_file=data_root + 'instances_trainval2017.json',
        img_prefix=data_root + 'images/',
        pipeline=train_pipeline),
    val=dict(
        classes=classes,
        type=dataset_type,
        ann_file=data_root + 'instances_val2017.json',
        img_prefix=data_root + 'images/',
        pipeline=test_pipeline),
    test=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017.json',
        img_prefix=data_root + 'val2017/',
        pipeline=test_pipeline))

# I put data_root here, because mixup may need this

optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
    policy='step',
    warmup='linear',
    warmup_iters=500,
    warmup_ratio=1.0 / 3,
    step=[8, 11])
checkpoint_config = dict(interval=1)
# yapf:disable
log_config = dict(
    interval=50,
    hooks=[
        dict(type='TextLoggerHook'),
        # dict(type='TensorboardLoggerHook')
    ])
# yapf:enable
# runtime settings
total_epochs = 12
dist_params = dict(backend='nccl')
log_level = 'INFO'
#  load_from = None
load_from = 'weights/cascade_rcnn_r50_fpn_dconv_c3-c5_1x_coco_20200130-2f1fca44.pth'
resume_from = None
workflow = [('train', 1)]
fp16 = dict(loss_scale=512.0)

看起来还行

提交

from mmdet.apis import init_detector, inference_detector
config_file = '/home/yons/workplace/python/mmdet/configs/underwater2021/cas_r50.py'
checkpoint_file = '/home/yons/workplace/python/mmdet/work_dirs/underwater2021/cas_r50/latest.pth'
model = init_detector(config_file, checkpoint_file, device='cuda:0')
import nncv
img_paths = nncv.scandir('/home/yons/data/underwater2021/test-A-image', 'jpg')
print(len(img_paths))
import glob
import numpy as np
results = []
from tqdm import tqdm
for img_path in tqdm(img_paths):
    img_name = os.path.basename(img_path)
    image_id = img_name.replace('.jpg', '')
    result = inference_detector(model, img_path)
    bboxes = np.vstack(result)
    labels = [
        np.full(bbox.shape[0], i, dtype=np.int32)
        for i, bbox in enumerate(result)
    ]
    labels = np.concatenate(labels)
    num = len(labels)
    for i in range(num):
        name = class_names[labels[i]]
        bbox = bboxes[i]
        xmin, ymin, xmax, ymax, confidence = bbox
        result = dict(name=name, image_id=image_id, confidence=confidence, xmin=xmin, ymin=ymin, xmax=xmax, ymax=ymax)
        results.append(result)
import pandas as pd
test_df = pd.DataFrame(results, columns=['name', 'image_id', 'confidence', 'xmin', 'ymin', 'xmax', 'ymax'])
test_df.to_csv('/home/yons/workplace/python/underwater2021/submit/0413.csv', index=False)

可视化看看

import random
img_path = random.choice(img_paths)
result = inference_detector(model, img_path)
class_names = ['holothurian', 'echinus', 'scallop', 'starfish']
from nncv.coco import vis_predict
vis_predict(img_path, result, class_names=class_names)