The ORBIT dataset is a collection of videos of objects in clean and cluttered scenes recorded by people who are blind/low-vision on a mobile phone. The dataset is presented with a teachable object recognition benchmark task which aims to drive few-shot learning on challenging real-world data.
microsoft
machine-learning
computer-vision
video
benchmark
dataset
classification
few-shot-learning
meta-learning
object-recognition
Обновлено 2024-08-13 03:27:45 +03:00
Lite Self-Training
Обновлено 2023-07-25 20:31:38 +03:00
Improving Generalization via Scalable Neighborhood Component Analysis
deep-learning
computer-vision
transfer-learning
visual-recognition
eccv-2018
few-shot-learning
nearest-neighbors
Обновлено 2023-06-12 22:02:13 +03:00