Efficient Pipeline for Camera Trap Image Review

Abstract

Biologists all over the world use camera traps to monitor biodiversity and wildlife population density. The computer vision community has been making strides towards automating the species classification challenge in camera traps, but it has proven difficult to to apply models trained in one region to images collected in different geographic areas. In some cases, accuracy falls off catastrophically in new region, due to both changes in background and the presence of previously-unseen species. We propose a pipeline that takes advantage of a pre-trained general animal detector and a smaller set of labeled images to train a classification model that can efficiently achieve accurate results in a new region.

From the Data Mining and AI for Conservation Workshop at KDD19

References in corpus (1)

• Camera-trap images segmentation using multi-layer robust principal component analysis

Cited by in corpus (10)

• Seeing biodiversity: perspectives in machine learning for wildlife conservation
• WILDS: A Benchmark of in-the-Wild Distribution Shifts
• Perspectives on individual animal identification from biology and computer vision
• Overcoming the Distance Estimation Bottleneck in Estimating Animal Abundance with Camera Traps
• ElephantBook: A Semi-Automated Human-in-the-Loop System for Elephant Re-Identification
• The iWildCam 2021 Competition Dataset
• A Bird Song Detector for improving bird identification through Deep Learning: a case study from Doñana
• Approximate Selection with Guarantees using Proxies
• The iWildCam 2020 Competition Dataset
• Image-to-Image Translation of Synthetic Samples for Rare Classes