Sara Beery, Stefan Schneider, and Jason Parham are hosting an Animal Re-Identification Workshop during the 2020 Winter Conference on Applications of Computer Vision (WACV 2020). This workshop will facilitate the formation of interdisciplinary connections and collaborations between the ecological and computer vision communities and start the discussion between researchers working with data across the image, audio, and video domains.
When: March 1st, 2020
Where: Snowmass Village, CO
Submission deadline: December 15th, 2019
We're excited to present our animal re-identification workshop for WACV2020. Animal re-identification is a technique used to answer many fundamental ecological hypotheses related to animal behavior and population dynamics. We believe the use of machine learning methods can aid and inspire fundamental research questions as well as open up the possibilities for hypotheses previously too difficult to test. We hope to keep our focus broad, but inclusive, so that everyone feels empowered to promote solutions to ecological questions. We highly encourage submissions which make their data publicly available
Topics that are anticipated, but are not required, include:
- Machine Learning for Human/Animal Re-Identification
- Vocal Recognition for Re-Identification
- Human/Animal Pose Estimation
- Behaviour Analysis from Audio/Video
- Data Augmentation for Limited Datasets
- Learning from Simulated Data
- Metric Learning
- One-Shot/ Zero-Shot Learning
- GPS Location Prediction
- Interpretability of AI Systems
- Public Ecological Datasets
- Best Practices for Data Collection and Curation
Call for Papers
We invite submissions of short papers using machine learning to address problems in image, video, and audio re-identification. We are interested in bringing the ecological and computer vision communities together, centered around research involving deep learning methods for animal re-identification applications. We would like to encourage contributors from the machine learning community to present cutting edge research relevant to animal re-identification, whether image, video, or audio processing. We would also encourage ecologists with limited to no machine learning expertise to come and discuss their ecological research problems in an attempt to brainstorm how machine learning may be of assistance. In addition, we welcome everyone in between.
We call for papers of three categories: Contributed Papers, Extended Abstracts, and Proposals
See the Submission Details page for more information
Sunday, March 1st, 2020
25 minutes long allowing 5 minutes for questions
Post-workshop socializing and brainstorming event for ideas related to the talks presented. All are welcome!
December 15th, 2019
January 20th, 2020
Camera Ready Papers Due
February 1st, 2020
For additional information on submission details, speakers, motivations, contact info, hosts, and more, please visit the Animal Re-ID Website.
Join the Machine and Deep Learning community group to connect with others working in this space.