Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

GradCam #43

Open
chrisdottel opened this issue Aug 16, 2022 · 1 comment
Open

GradCam #43

chrisdottel opened this issue Aug 16, 2022 · 1 comment

Comments

@chrisdottel
Copy link

I'd love to try and implement GradCam on this project, however the pretrained models seem to be packed with the Tensorflow save instead of Keras. Do you have the checkpoints available for the evaluation or training code for finetuning/ custom implementations?
@isarandi
Copy link
Owner

Do you necessarily need a Keras-style checkpoint for this? It should be possible to backprop through the TF SavedModel as well and get gradients wrt the input image. You may want to use the model.crop_model for this analysis, which is the single-person pose estimator at the core of the method, which takes a 256x256 cropped image. (See bottom of https://github.com/isarandi/metrabs/blob/master/docs/INFERENCE_GUIDE.md)

Young-Yoon pushed a commit to Young-Yoon/metrabs that referenced this issue Mar 8, 2024
@inakinavarro
Add automatic model conversion to onnx, ncnn and ave-tracker bundle (i…
d12e02c 
…sarandi#43)

This PR brings the functionality for model conversion inspired by the avelab model conversion pipeline used in face tracking.

The converter takes as input the model folder with pb files and variable folder from a training and converts it into several onnx files. If ave-tracker path is given then the models are converted into ncnn files (optimized and binarized), a bundle file to be used in ave-tracker containing a settings.json and the ncnn file is also generated. The settings.json file is automatically created. Furthermore a info.txt file is created with md5 checksum of the relevant files used as inputs.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@isarandi @chrisdottel