HOD: A Benchmark Dataset for Harmful Object Detection.
Eungyeom HaHeemook KimSung Chul HongDongbin NaPublished in: CoRR (2023)
Keyphrases
- benchmark datasets
- object detection
- pedestrian detection
- face detection
- object categories
- computer vision
- object segmentation
- object recognition
- object classification
- ensemble methods
- background subtraction
- terms of classification accuracy
- scene understanding
- object class
- multi class
- discriminatively trained
- human detection
- pascal voc
- scene recognition
- deformable part models
- genetic algorithm
- human body
- image representation
- support vector machine
- classification accuracy
- viewpoint
- expert systems