The PanAf-SBR Dataset: Social Behaviour Recognition for Wild Great Apes

2026-07-19

AI Summary

This study addresses the challenge of automatically recognizing fine-grained social behaviors of wild great apes, which has been hindered by the absence of annotated camera-trap datasets. To bridge this gap, we introduce PanAf-SBR, the first dataset of its kind, comprising 36,063 images with 81,096 fine-grained annotations—including bounding boxes, segmentation masks, individual identities, and seven categories of social behaviors—annotated using an actor–recipient paradigm. Building upon the AlphaChimp architecture, we establish a benchmark model enhanced by bidirectional transfer learning and a mask-inversion background suppression technique. Experiments demonstrate that cross-dataset pretraining substantially improves performance on specific behavior recognition tasks, and that contextual background information plays a critical role in behavior understanding. This work provides a novel tool for monitoring great ape social structures and enabling early warnings of population decline.

Abstract

Behavioural shifts in wild great ape populations, particularly the breakdown of social structures, can serve as an early indicator of population decline. Automating the detection of behaviours indicative of these shifts is therefore a critical task for conservation. Several valuable datasets have recently been introduced for the automated recognition of great ape behaviour, yet few include fine-grained social behaviour annotations, and those that do are captured either in captive settings or via aerial platforms such as UAVs. We address this gap by introducing PanAf-SBR, the first wild great ape camera trap dataset annotated with social behaviours. PanAf-SBR extends PanAf500 with 100 additional videos covering 36,063 frames. These come with 81,096 annotations including bounding boxes, segmentation masks, intra-video identities, and seven social behaviour classes defined under the action giver and receiver convention of ChimpACT. We use this data together with the AlphaChimp architecture to establish the first benchmarks for fine-grained social behaviour recognition in wild great apes from camera trap footage. We further conduct bidirectional transfer learning experiments between PanAf-SBR and the captive ChimpACT dataset, finding that cross-dataset pre-training is highly beneficial for specific classes rather than of uniform benefit. Finally, we examine the role of background context by inverting the segmentation masks to suppress non-ape pixels.

Problem

• social behaviour recognition
• wild great apes
• camera trap dataset
• behavioural annotation
• conservation monitoring

Innovation

• social behaviour recognition
• wild great apes
• camera trap dataset
• fine-grained annotation
• transfer learning