HQA-Trans: An end-to-end high-quality-awareness image translation framework for unsupervised cross-domain pedestrian detection.
Gelin ShenYang YuZhi-Ri TangHaoqiang ChenZongtan ZhouPublished in: IET Comput. Vis. (2022)
Keyphrases
- end to end
- cross domain
- pedestrian detection
- high quality
- human body
- image features
- input image
- multiscale
- image classification
- object detection
- image retrieval
- high resolution
- semi supervised
- histograms of oriented gradients
- congestion control
- text categorization
- bit rate
- prior knowledge
- knowledge transfer
- similarity measure
- feature selection
- real world