AnimalMotionCLIP: Embedding motion in CLIP for Animal Behavior Analysis

TL;DR

AnimalMotionCLIP enhances animal behavior recognition by integrating motion cues and temporal modeling into the CLIP framework, leading to improved accuracy in recognizing fine temporal actions.

Contribution

This work introduces a novel method that incorporates motion information and various temporal aggregation schemes into CLIP for better animal behavior analysis.

Findings

Outperforms state-of-the-art methods on Animal Kingdom dataset

Effective integration of optical flow improves behavior recognition

Temporal modeling schemes enhance recognition of fine actions

Abstract

Recently, there has been a surge of interest in applying deep learning techniques to animal behavior recognition, particularly leveraging pre-trained visual language models, such as CLIP, due to their remarkable generalization capacity across various downstream tasks. However, adapting these models to the specific domain of animal behavior recognition presents two significant challenges: integrating motion information and devising an effective temporal modeling scheme. In this paper, we propose AnimalMotionCLIP to address these challenges by interleaving video frames and optical flow information in the CLIP framework. Additionally, several temporal modeling schemes using an aggregation of classifiers are proposed and compared: dense, semi dense, and sparse. As a result, fine temporal actions can be correctly recognized, which is of vital importance in animal behavior analysis. Experiments on the Animal Kingdom dataset demonstrate that AnimalMotionCLIP achieves superior performance compared to state-of-the-art approaches.