One-shot action recognition in challenging therapy scenarios

TL;DR

This paper introduces a robust one-shot action recognition method for challenging scenarios, including therapy with autistic individuals, outperforming existing models and providing valuable real-time evaluation metrics.

Contribution

A novel motion representation approach for one-shot action recognition in complex, real-world settings, including therapy scenarios with high artifacts.

Findings

Outperforms previous models on NTU-120 benchmark

Effective in challenging therapy scenarios with high motion artifacts

Provides real-time qualitative and quantitative assessment

Abstract

One-shot action recognition aims to recognize new action categories from a single reference example, typically referred to as the anchor example. This work presents a novel approach for one-shot action recognition in the wild that computes motion representations robust to variable kinematic conditions. One-shot action recognition is then performed by evaluating anchor and target motion representations. We also develop a set of complementary steps that boost the action recognition performance in the most challenging scenarios. Our approach is evaluated on the public NTU-120 one-shot action recognition benchmark, outperforming previous action recognition models. Besides, we evaluate our framework on a real use-case of therapy with autistic people. These recordings are particularly challenging due to high-level artifacts from the patient motion. Our results provide not only quantitative but also online qualitative measures, essential for the patient evaluation and monitoring during the actual therapy.