ProfVLM: A Lightweight Video-Language Model for Multi-View Proficiency Estimation
This addresses the problem of explainable skill assessment for domains like education or training, offering a novel generative approach that is more efficient and transparent than existing methods.
The paper tackles skill proficiency estimation by introducing ProfVLM, a lightweight video-language model that predicts skill levels and generates expert feedback from multi-view videos, achieving superior accuracy with up to 20x fewer parameters and 60% faster training.
Existing approaches to skill proficiency estimation often rely on black-box video classifiers, ignoring multi-view context and lacking explainability. We present ProfVLM, a compact vision-language model that reformulates this task as generative reasoning: it jointly predicts skill level and generates expert-like feedback from egocentric and exocentric videos. Central to our method is an AttentiveGatedProjector that dynamically fuses multi-view features, projected from a frozen TimeSformer backbone into a language model tuned for feedback generation. Trained on EgoExo4D with expert commentaries, ProfVLM surpasses state-of-the-art methods while using up to 20x fewer parameters and reducing training time by up to 60%. Our approach not only achieves superior accuracy across diverse activities, but also outputs natural language critiques aligned with performance, offering transparent reasoning. These results highlight generative vision-language modeling as a powerful new direction for skill assessment.