Toward Annotation Efficient Learning for Computer Vision
|2020||Facebook AI Research||방문연구원|
Recent progress in machine learning has led to many advances in engineering and science fields, including computer vision and graphics. Most notable successes have seen in supervised learning with deep neural networks. Despite these successes, as a consequence of using the high capacity models, we are struggling with a lack of high-quality annotations in supervised methods. Obtaining a massive amount of carefully annotated and curated data is often expensive or even challenging.
In this talk, I will present a few potential strategies for case-study, including self-/semi-supervised learning, generating synthetic data, injeting a prior to an architecture, borrowing different domain data, etc.
ID : 897 821 7407
PW : 1nTQDY