Questions Regarding AlphaPose Implementation and Output Interpretation

[ignacio-barrera]

I am currently working on a project that involves testing some Human Pose Estimation Models, my principal focus its to track foots, so im using AlphaPose, OpenPose and BlazePose, because this 3 models ahs 3 foot keypoints each. So, my plan its to test this three models on COCO-WholeBody, which has 133 keypoints (3 for each feet).
I have encountered a few questions and problems that would greatly appreciate your guidance, my main three problems and questions are:

• Clarification on Keypoints Indices: I understand that AlphaPose outputs 133 keypoints for whole body pose estimation. Could you provide a detailed description that clearly maps each keypoint index to its corresponding body part? Specifically, I am looking to identify keypoints related to the feet, additionally, is there an example or best practice for isolating and processing these specific keypoints? My problems its that the output its a array of a lot of numbers and they not have any type of identification so i dont know how to use it. I used this command for inference " !python scripts/demo_inference.py --cfg configs/coco_wholebody/resnet/256x192_res50_lr1e-3_2x-regression.yaml --checkpoint pretrained_models/coco_wholebody133_fast50_regression_256x192.pth --indir ../../dataset/val2017/ --outdir examples/res_wholebody_val2017 --eval --vis_fast --save_img ".

• Script for validate: I also tried your script validate.py [1] for validate on WholeBody (eval2017), but i have an error with a dismatch of shaps (17 != 133) purposes, I am using this commnand " !python scripts/validate.py --cfg configs/coco_wholebody/resnet/256x192_res50_lr1e-3_2x-regression.yaml --checkpoint pretrained_models/coco_wholebody133_fast50_regression_256x192.pth ", I dont understand why becausa i read de config file and everything its for 133 keypoints so i dont know what to do.

• Memory Error Handling: While processing a large number of images (5000 img, COCO eval2017), I occasionally encounter memory allocation errors (e.g., numpy.core._exceptions._ArrayMemoryError). Could you provide advice on how the model use memory? Why is the error? I mean the models use the RAM memory for that 5k images? Or how it this works?.

All of this its using FastPose Model for COCO WholeBody dataset 133 keypoints [2] (checkpoint named coco_wholebody133_fast50_regression_256x192.pth) and this cfg file [3] (cfg file named: 256x192_res50_lr1e-3_2x-regression.yaml).

And last, the output for my inference (COCO eval2017) looks like the .json file that i adjunted.

I greatly appreciate your time and assistance in addressing these questions.

Links:
https://github.com/MVIG-SJTU/AlphaPose/blob/master/scripts/validate.py
https://github.com/MVIG-SJTU/AlphaPose/blob/master/docs/MODEL_ZOO.md
https://github.com/MVIG-SJTU/AlphaPose/blob/master/configs/coco_wholebody/resnet/256x192_res50_lr1e-3_2x-regression.yaml

first_res.json