Egor Beliaev

Results 10 issues of Egor Beliaev

Histograms of the ground truth and predicted scores from the article, p.7 ![original](https://user-images.githubusercontent.com/2828322/50113240-0cf92200-024a-11e9-8e70-e663460bc6a8.png) And did the same for two models here: ![titu1994](https://user-images.githubusercontent.com/2828322/50113285-269a6980-024a-11e9-8c92-5aed8e50dcf7.png) It shows that MobileNet will show correct scores...

This is the loop for your stage1 which invokes first model repeatedly: ``` def stage1(img, imgPad, net, thres): row = (imgPad.shape[0] - img.shape[0]) // 2 col = (imgPad.shape[1] - img.shape[1])...

For 418x449 image I have `[600,644]` written to the features, and for 1042x480 I have `[1000x461]`. Those are not the values for some other images from my dataset, as I...

[https://drive.google.com/file/d/1HA7mlvird_xtzrd-CdqKRykFHsR4PgDp/view?usp=sharing](https://drive.google.com/file/d/1HA7mlvird_xtzrd-CdqKRykFHsR4PgDp/view?usp=sharing)

Hi, I wanted to share the way to help with this problem: > Training seems to suffer from mode collapse (when trained wrong, enhanced image looks like negatively processed image)....

Compilation yields errors which I believe are due to incompatibility with OpenCV 3 (e.g. `‘class cv::Mat’ has no member named ‘refcount`): ``` /home/hcl/Documents/study/opencv/deepcut-cnn/python/pose/cpp/numpy-opencv-converter-master/utils/conversion.cpp:188:16: error: cannot declare variable ‘g_numpyAllocator’ to be...

[https://drive.google.com/file/d/1ho84S78NxuwhAeEAJY11B5v6PUBnQkKo/view?usp=sharing](https://drive.google.com/file/d/1ho84S78NxuwhAeEAJY11B5v6PUBnQkKo/view?usp=sharing)

[https://drive.google.com/file/d/1ho84S78NxuwhAeEAJY11B5v6PUBnQkKo/view?usp=sharing](https://drive.google.com/file/d/1ho84S78NxuwhAeEAJY11B5v6PUBnQkKo/view?usp=sharing)

Hi, the segmentation seems better than commonly used Bisenet, espectially it is not limited to crop - which is great. But I see artifacs (wide vertical line) on some images...

Pytorch 1.3 used for training (Ok) and for inference. I had a problem with Dataparallel, which I solved by copying multi-gpu model load code in into 1 gpu case: ```...