FDLNet: Boosting Real-time Semantic Segmentation by Image-size Convolution via Frequency Domain Learning.
Qingqing YanShu LiChengju LiuMing LiuQijun ChenPublished in: ICRA (2023)
Keyphrases
- frequency domain
- semantic segmentation
- discrete fourier transform
- fourier domain
- spatial domain
- fourier transform
- weakly supervised
- fast fourier transform
- learning algorithm
- image data
- image analysis
- segmentation method
- image features
- input image
- image representation
- single image
- multiscale
- multiresolution
- image regions
- spatial information
- image classification
- scene classification
- object detectors
- conditional random fields
- region of interest
- image registration
- test images
- superpixels
- pixel level
- similarity measure
- high resolution
- computer vision
- object recognition
- computational complexity
- subband