Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2 | bioRxiv

Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2 | bioRxiv

Motion-activated wildlife cameras (or 'camera traps') are frequently used to remotely and non-invasively observe animals. The vast number of images collected from camera trap projects have prompted some biologists to employ machine learning algorithms to automatically recognize species in these images, or at least filter-out images that do not contain animals. These approaches are often limited by model transferability, as a model trained to recognize species from one location might not work as well for the same species in different locations. Furthermore, these methods often require advanced computational skills, making them inaccessible to many biologists. We used 3 million camera trap images from 18 studies in 10 states across the United States of America to train two deep neural networks, one that recognizes 58 species, the 'species model,' and one that determines if an image is empty or if it contains an animal, the 'empty-animal model.' Our species model and empty-animal model ha

2 mentions: @Mikey_QSC@jeffclune
Date: 2020/03/24 15:51

Referring Tweets

@Mikey_QSC MLWIC2: R Shiny Apps to use and train #MachineLearning models for #CameraTrap images. We provide 2 trained models with 97% accuracy: 1 Recognizes 58 species 2 Filters out empty images If these don’t work, train your own model in a point and click GUI t.co/C0oeADt2Xe t.co/q1NLMxQGgC
@jeffclune New paper harnessing deep learning for environmental conservation, to combat poaching, and improve ecological understanding. New model recognizes 58 species with 97% accuracy, & filters out empty images, + a point and click GUI. t.co/xaWo0eNb93 Led by M. Tabak @Mikey_QSC t.co/wL1fYkv66D

Related Entries

Read more Montezuma's Revenge Solved by Go-Explore, a New Algorithm for Hard-Exploration Problems (Sets Record...
0 users, 36 mentions 2018/12/09 04:30
Read more ICML 2019 Tutorial: Recent Advances in Population-Based Search for Deep Neural Networks - YouTube
0 users, 1 mentions 2019/10/11 18:48
Read more Variational autoencoder for Lego faces - echevarria.io
0 users, 5 mentions 2020/02/06 09:50
Read more [2003.11164] Fiber: A Platform for Efficient Development and Distributed Training for Reinforcement ...
0 users, 1 mentions 2020/03/26 15:50
Read more Introduction to Fiber - Fiber
0 users, 3 mentions 2020/03/26 15:50