Husan

@Geo-fortune. I removed my opencv and reinstall it again and worked for me.

@jiezhang165 Follow the guidance of this link: https://github.com/PaddlePaddle/Research/tree/master/CV/PaddleReid/process_aicity_data You have to prepare your dataset before training. You may also need to make some other changes. 遵循此链接的指导：https://github.com/PaddlePaddle/Research/tree/master/CV/PaddleReid/process_aicity_data 您必须准备要训练的数据集。 您可能还需要做一些其他更改。 > 请问训练和测试用的数据集是AICity20...

@layumi @s20163081161 I am running test_2020.py and as @layumi has mentioned earlier I am going to use `val` for test. But how could I use `val` as a test dataset?...

@layumi Thank you for your support. I have split val set to `gallery` and `query` as following if it is right way: val/000341/ -- 000341_c010_7.jpg -- 000341_c016_11.jpg -- 000341_c017_5.jpg query/000341/...

@layumi 您好 When I crop images, should I crops images from `val` set or directly from AIC20_ReID's `image_query` and `image_test` sets. I am really confused about what dataset should I...

> You may run `prepare_cam2020.py` to prepare your data, which will put the images of the same camera in one subfolder. > Then you can run the classification model to...

Yes, it looks very slow. It is taking 1.2 second for per image with RTX 2080 when I run test.py file.

Ok. I found the reason. My `nframes` is equal to `inf`. I don't know why. ```python {'plugin': 'ffmpeg', 'nframes': inf, 'ffmpeg_version': '4.2.2-static https://johnvansickle.com/ffmpeg/ built with gcc 8 (Debian 8.3.0-6)', 'codec':...

Ok. I have solved it. Instead of `n_frames = meta['nframes']` I checked for its type whether it is string or not. if it is string then you can do as...

Did anyone solve the problem?