Towards Zero-Shot Camera Trap Image Categorization
This addresses the problem of location-specific overfitting in camera trap image analysis for ecological monitoring, though it is incremental with hybrid methods.
The paper tackled automatic categorization of camera trap images by benchmarking and combining methods like MegaDetector with classifiers, reducing relative error by up to 75% on datasets, and proposing a zero-shot pipeline that achieved competitive results.
This paper describes the search for an alternative approach to the automatic categorization of camera trap images. First, we benchmark state-of-the-art classifiers using a single model for all images. Next, we evaluate methods combining MegaDetector with one or more classifiers and Segment Anything to assess their impact on reducing location-specific overfitting. Last, we propose and test two approaches using large language and foundational models, such as DINOv2, BioCLIP, BLIP, and ChatGPT, in a zero-shot scenario. Evaluation carried out on two publicly available datasets (WCT from New Zealand, CCT20 from the Southwestern US) and a private dataset (CEF from Central Europe) revealed that combining MegaDetector with two separate classifiers achieves the highest accuracy. This approach reduced the relative error of a single BEiTV2 classifier by approximately 42\% on CCT20, 48\% on CEF, and 75\% on WCT. Besides, as the background is removed, the error in terms of accuracy in new locations is reduced to half. The proposed zero-shot pipeline based on DINOv2 and FAISS achieved competitive results (1.0\% and 4.7\% smaller on CCT20, and CEF, respectively), which highlights the potential of zero-shot approaches for camera trap image categorization.