ChocoTiamo

> Hello @tiamo405 ! you can use OpenPose with configuration BODY_25 (25 keypoints) to extract the pose, instead of coco (18 keypoints).Here there's a photo of both configurations: https://www.researchgate.net/figure/Architecture-of-the-different-modules-to-recognise-poses-and-control-the-robot_fig3_340618434 thank...

> Original imagees can be downloaded in [Google Drive](https://drive.google.com/file/d/1ZirYG7n-9QL0G1Ub8bkDUKRMI7HwrvQv/view?usp=sharing). More details about dataset can be found in [DATASET.md](https://github.com/yl-1993/learn-to-cluster/blob/master/DATASET.md). I see the file data/features/part1_test.bin . In this file are the features...

> Also, it may be noticed that the dropbox storing the re-aligned face images https://www.dropbox.com/s/wpx6tqjf0y5mf6r/faces_ms1m-refine-v2_112x112.zip?dl=0 cannot be accessed. Could you please help to take a look if available? Thanks. hi,...

mn cho mình hỏi sao mình k tìm thấy other settings ạ, thiếu mỗi chỗ tích ở đó với lỗi java: java.lang.ClassNotFoundException: org.hibernate.jpamodelgen.JPAMetaModelEntityProcessor ạ

I use pretrained : pretrained/adaface_ir50_ms1mv2.ckpt, can I use a different image from the data train for inference?

i tried inference 5 images and the return result is not very good. 2 identical pictures, the girl is worth 1, but with 2 pictures of Tony Stark is not...

Have you found the reason why? If left as is, can it be trained?

line 207 train.py: pitch, yaw = model(images_gaze) line 22,23: self.fc_yaw_gaze = nn.Linear(512 * block.expansion, num_bins) self.fc_pitch_gaze = nn.Linear(512 * block.expansion, num_bins) line 68 69 70: pre_yaw_gaze = self.fc_yaw_gaze(x) pre_pitch_gaze =...

I tried to download the dataset but it seems like the dataset is too big, can anyone split the dataset for me around 5gb so I can download it?

> @tiamo405 , MPIIFaceGaze is relatively smaller than Gaze360. MPIIFaceGaze not folder Label, link download: https://www.mpi-inf.mpg.de/departments/computer-vision-and-machine-learning/research/gaze-based-human-computer-interaction/its-written-all-over-your-face-full-face-appearance-based-gaze-estimation