PedVis-VGG-16: A Fine-tuned deep convolutional neural network for pedestrian image classifications.
Mahassine DefaouiLahcen KouttiMohamed El AnsariRedouan LahmyedLhoussaine MasmoudiPublished in: WINCOM (2022)
Keyphrases
- image features
- single image
- input image
- image data
- multiscale
- convolutional neural network
- image content
- image retrieval
- grey level
- template matching
- fine tuned
- region of interest
- image representation
- image regions
- image analysis
- image classification
- feature points
- high resolution
- test images
- keypoints
- pedestrian detection
- lighting conditions
- low level
- segmentation method
- energy function
- segmentation algorithm
- edge detection
- image database