Domain-Adaptive Pretraining Improves Primate Behavior Recognition
This addresses the bottleneck of data efficiency for researchers in ecology and conservation by enabling better recognition of animal behavior without labeled samples.
The paper tackles the problem of high labeling costs in primate behavior recognition from video camera traps by using domain-adaptive pretraining with self-supervised learning, achieving improvements of 6.1% accuracy and 6.3% mAP over state-of-the-art models on two datasets.
Computer vision for animal behavior offers promising tools to aid research in ecology, cognition, and to support conservation efforts. Video camera traps allow for large-scale data collection, but high labeling costs remain a bottleneck to creating large-scale datasets. We thus need data-efficient learning approaches. In this work, we show that we can utilize self-supervised learning to considerably improve action recognition on primate behavior. On two datasets of great ape behavior (PanAf and ChimpACT), we outperform published state-of-the-art action recognition models by 6.1 %pt. accuracy and 6.3 %pt. mAP, respectively. We achieve this by utilizing a pretrained V-JEPA model and applying domain-adaptive pretraining (DAP), i.e. continuing the pretraining with in-domain data. We show that most of the performance gain stems from the DAP. Our method promises great potential for improving the recognition of animal behavior, as DAP does not require labeled samples. Code is available at https://github.com/ecker-lab/dap-behavior