CarryHJR

modelzoo里面的mAP val指标 看起来 和 AP50指标弄混了

1

https://github.com/wuzhihao7788/yolodet-pytorch/blob/master/docs/MODEL_ZOO_cn.md 比如 网络 | 预训练数据集 | 输入尺寸 | epoch | 显卡类型 | 推理时间(fps) | cocotools APval | AP50 | 百度网盘 | 谷歌网盘 | 配置文件 | 日志 -- | -- |...

``` ./tools/dist_train.sh configs/LogDet/swinb_3x_bigsize_anchor_bs2x8.py 8 ./tools/dist_test.sh configs/LogDet/swinb_3x_bigsize_anchor_bs2x8.py work_dirs/LogDet/swinb_3x_bigsize_anchor_bs2x8/latest.pth 8 --format-only --options "jsonfile_prefix=work_dirs/LogDet/swinb_3x_bigsize_anchor_bs2x8" ```

这里贴一下划分val的代码 ``` coco = COCO(os.path.join(root, 'annotations/instances_train2017.json')) class_names = [coco.cats[catId]['name'] for catId in coco.getCatIds()] categories = [dict(id=i+1, name=name) for i, name in enumerate(class_names)] annotaions_train = [] images_train = [] annotaions_val =...

the official detectron support cascade r50 fpn https://dl.fbaipublicfiles.com/detectron2/COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x/137849600/model_final_f10217.pkl only 170 mb but in your implementation, the model size is about 570mb

变化检测: 链接: https://pan.baidu.com/s/1zt8KHwtuVFkXSoJTB53ryA 提取码: rw4f 场景分类 链接: https://pan.baidu.com/s/1RoY96YqFZm2drzqyd8RL5Q 提取码: xb4r