YouTube-BoundingBoxes: A Large High-Precision Human-Annotated Data Set for Object Detection in Video.
Esteban RealJonathon ShlensStefano MazzocchiXin PanVincent VanhouckePublished in: CoRR (2017)
Keyphrases
- high precision
- object detection
- data sets
- high recall
- high reliability
- video search
- multimedia
- video sharing
- high accuracy
- video sequences
- action recognition
- video dataset
- human activities
- achieve high precision
- video database
- youtube videos
- action detection
- static images
- human actions
- social media
- face detection
- space time
- video streams
- video clips
- video content
- multimedia data
- online video
- spatio temporal
- video retrieval
- video analysis
- motion capture data
- training data