Performance of Large Pre-trained Gesture Recognition Models on Real-world Datasets after Synthetic Fine-tuning

In this work, we explore the possibility of using synthetically generated data for video-based gesture recognition with large pre-trained models. We consider whether these models have sufficiently robust and expressive representation spaces to enable "training-free" classification. Specifically, we utilize various state-of-the-art video encoders to extract features for use in k-nearest neighbors classification, where the training data points are derived from synthetic videos only. We compare these results with another training-free approach -- zero-shot classification using text descriptions o

Research goal: How do large pre-trained models for gesture recognition perform when evaluated on real-world video datasets after fine-tuning on synthetically generated videos, measured by classification accuracy and generalization to unseen gestures?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.6/10.