Human Pose Descriptions and Subject-Focused Attention for Improved Zero-Shot Transfer in Human-Centric Classification Tasks

TL;DR

This paper introduces a new approach combining human pose descriptions and subject-focused attention in CLIP to improve zero-shot human-centric classification, demonstrating significant accuracy gains across multiple datasets and tasks.

Contribution

It presents a novel dataset of natural language pose descriptions and a new FocusCLIP framework with Subject-Focused Attention for enhanced zero-shot classification performance.

Findings

8.61% average accuracy improvement over CLIP

Significant gains in activity, age, and emotion recognition

Effective use of pose descriptions for zero-shot transfer

Abstract

We present a novel LLM-based pipeline for creating contextual descriptions of human body poses in images using only auxiliary attributes. This approach facilitates the creation of the MPII Pose Descriptions dataset, which includes natural language annotations for 17,367 images containing people engaged in 410 distinct activities. We demonstrate the effectiveness of our pose descriptions in enabling zero-shot human-centric classification using CLIP. Moreover, we introduce the FocusCLIP framework, which incorporates Subject-Focused Attention (SFA) in CLIP for improved text-to-image alignment. Our models were pretrained on the MPII Pose Descriptions dataset and their zero-shot performance was evaluated on five unseen datasets covering three tasks. FocusCLIP outperformed the baseline CLIP model, achieving an average accuracy increase of 8.61\% (33.65\% compared to CLIP's 25.04\%). Notably, our approach yielded improvements of 3.98\% in activity recognition, 14.78\% in age classification, and 7.06\% in emotion recognition. These results highlight the potential of integrating detailed pose descriptions and subject-level guidance into general pretraining frameworks for enhanced performance in downstream tasks.