Improved YOLOv7 models based on modulated deformable convolution and swin transformer for object detection in fisheye images.
Jie ZhouDegang YangTingting SongYichen YeXin ZhangYingze SongPublished in: Image Vis. Comput. (2024)
Keyphrases
- object detection
- image database
- image data
- ground truth
- three dimensional
- input image
- random fields
- image retrieval
- image registration
- edge detection
- image analysis
- linear filtering
- face detection
- object recognition
- image classification
- test images
- background subtraction
- parametric models
- object detectors
- fourier domain
- discrete fourier transform
- probabilistic model
- category level