Abstract

Along with the exponential growth of online video creation platforms such as TikTok and Instagram, state of the art research involving quick and effective human action/gesture recognition remains crucial. This thesis presents an instance-based statistical framework which addresses the challenge of classifying short human action video clips, using a domain-specific feature design approach, capable of performing significantly well using as little as one training example per action (one-shot learning). The method is based on Gunner Farneback's dense optical flow (GF-OF) estimation strategy, Gaussian mixture models, and information divergence. We first aim to obtain accurate representations of the human movements/actions by clustering the results given by GF-OF using K-means method of vector quantization. We then proceed by representing the result of one instance of each action by a Gaussian mixture model. Furthermore, using Kullback-Leibler divergence (KL-divergence), we estimate the information divergences in an attempt to find similarities between the trained actions and the ones in the test videos. Classification is then done by matching each test video to the trained action with the highest similarity (a.k.a lowest KL-divergence value). We have performed experiments on the KTH and Weizmann Human Action datasets using One-Shot and K-Shot learning approaches, and the results reveal the discriminative nature of our proposed methodology in comparison with state-of-the-art techniques.

Available Versions of this Item

• An Instance-Based Learning Statistical Framework for One-Shot and Few-Shot Human Action Recognition. (deposited 21 Feb 2022 17:27) [Currently Displayed]
  • An Instance-Based Learning Statistical Framework for One-Shot and Few-Shot Human Action Recognition. (deposited 16 Jun 2022 14:41)