19类实验记录
一.实验简介
1.实验索引
实验代码:https://github.com/salmanmaq/segmentationNetworks
实验数据集:https://www.kaggle.com/salmanmaq/m2caiseg
2.实验介绍
本次实验通过segmentationNetworks网络训练m2caiseg数据集,并统计了IoU,Precision,Recall,F1四个评估指标。
3.实验环境
本次实验使用kaggle notebook,默认torch版本为1.4。
二.实验内容
1.实验设计
由于存在torch版本不同的问题,根据shell脚本中参数配置,我对代码做出了本分修改。
'''
Training for combined laparoscopic image segmentation and tool presense
classification
'''
import argparse
import os
import shutil
import numpy as np
import cv2
from PIL import Image
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.backends.cudnn as cudnn
import torch.optim as optim
from torch.autograd import Variable
import torch.utils.data
import torchvision.transforms as transforms
import torch.nn.functional as F
import utils
from model.segnetPlusClass import segnetPlusClass
from datasets.miccaiSegPlusClassDataLoader import miccaiSegPlusClassDataset
from torchvision.transforms import InterpolationMode
parser = argparse.ArgumentParser(description='PyTorch SegNet Training')
parser.add_argument('-j', '--workers', default=4, type=int, metavar='N',
help='number of data loading workers (default: 4)')
parser.add_argument('--epochs', default=500, type=int, metavar='N',
help='number of total epochs to run')
parser.add_argument('--start-epoch', default=0, type=int, metavar='N',
help='manual epoch number (useful on restarts)')
parser.add_argument('--batchSize', default=2, type=int,
help='Mini-batch size (default: 4)')
parser.add_argument('--lr', '--learning-rate', default=0.005, type=float,
metavar='LR', help='initial learning rate')
parser.add_argument('--wd', '--weight_dacay', default=0.0005, type=float,
help='initial learning rate')
parser.add_argument('--bnMomentum', default=0.1, type=float,
help='Batch Norm Momentum (default: 0.1)')
parser.add_argument('--imageSize', default=256, type=int,
help='height/width of the input image to the network')
parser.add_argument('--resizedImageSize', default=224, type=int,
help='height/width of the resized image to the network')
parser.add_argument('--print-freq', '-p', default=1, type=int, metavar='N',
help='print frequency (default:1)')
parser.add_argument('--resume', default='/kaggle/input/checkpoint485/checkpoint_485.tar', type=str, metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('-e', '--evaluate', dest='evaluate', action='store_true',
help='evaluate model on