Hi Mario and Andrea,
My teammates Omar, Jack, and I are currently working on a research project at Monash University focused on the diagnostic analysis of facial emotion recognition (FER) models.
We have been exploring your vfer repository and were particularly interested in the VitFER implementation described in your ViT-Emotion-Recognition.ipynb notebook. Your approach to fine-tuning vision transformers for real-time emotion classification is very relevant to our current evaluation of transformer-based architectures.
We are currently setting up our environment in Google Colab to test the model's performance within our specific framework. However, we noticed that the fine-tuned model weights (checkpoints) do not appear to be directly available in the public repository.
Would you be able to provide access to the pre-trained weights for the VitFER model? Having access to the trained model would be incredibly helpful for our comparative analysis and would allow us to build upon your work more effectively.
Thank you for your time and for sharing your work with the community.
Kind regards,
Joshua Chan
Hi Mario and Andrea,
My teammates Omar, Jack, and I are currently working on a research project at Monash University focused on the diagnostic analysis of facial emotion recognition (FER) models.
We have been exploring your vfer repository and were particularly interested in the VitFER implementation described in your ViT-Emotion-Recognition.ipynb notebook. Your approach to fine-tuning vision transformers for real-time emotion classification is very relevant to our current evaluation of transformer-based architectures.
We are currently setting up our environment in Google Colab to test the model's performance within our specific framework. However, we noticed that the fine-tuned model weights (checkpoints) do not appear to be directly available in the public repository.
Would you be able to provide access to the pre-trained weights for the VitFER model? Having access to the trained model would be incredibly helpful for our comparative analysis and would allow us to build upon your work more effectively.
Thank you for your time and for sharing your work with the community.
Kind regards,
Joshua Chan