Georgios Pavlakos
Georgios Pavlakos
Please take a look at the answers [here](https://github.com/vchoutas/smplify-x/issues/3). Other users having the same problem were able to resolve it by following the instructions posted in one of the answers.
If you check the first log, the image that generates the NaN error is HICO_test2015_00000471.jpg. The other images (e.g., HICO_test2015_00000470.jpg) run without problems even when you have all the images...
I reproduced the procedure for downloading the test images and running the model on them, and I could indeed get the expected results on the full test set of h36m...
Given a bounding box of a person with width W, and height H, the variable scale is defined as max(W,H)/200. Also, the variable 'res' defines the resolution of the output...
If you want to run the demo on single images, the whole pipeline, including detection and reconstruction, will take 9.5Gb of GPU memory (this can go down to 5.2Gb if...
For track.py, the code also calls the [PHALP tracking code](https://github.com/brjathu/PHALP), which uses [egl](https://github.com/brjathu/PHALP/blob/8f5913f41466c54cf57a257b480343b4409cbf7f/phalp/visualize/py_renderer.py#L3), so, it's likely you get a conflict there? Is it possible for you to fix the egl...
When you run the [demo](https://github.com/shubham-goel/4D-Humans/blob/244be5a76dbf70f3937e7575de8bbc9c02ae6548/demo.py#L91), check the output of `out['pred_keypoints_3d']`. The names of the joints are listed [here](https://github.com/nkolot/SPIN/blob/2476c436013055be5cb3905e4e4ecfa86966fac3/constants.py#L16) (we return the first 44 of these 49 joints). If you want...
Not sure of the exact setting you are working with, but one aspect that we observed that could create issues with GPU memory is using values for the [number of...
When you set `trainer.strategy=ddp_find_unused_parameters_true`, do you actually get an error message, or only this warning? If you are only getting a warning, but the code runs, it should be ok.
When you say "you need to make the batch size _much smaller_", what do you compare it with? On our end, we are currently running the code with this additional...