A pose-based hand image classification method retrainable on embedded target devices.
Alessio CanepaEdoardo RagusaChristian GianoglioPaolo GastaldoRodolfo ZuninoPublished in: ICECS (2021)
Keyphrases
- classification method
- target images
- input image
- image retrieval
- position and orientation
- image features
- multiscale
- target object
- embedded systems
- image classification
- text classification
- classification scheme
- image segmentation
- k nearest neighbor
- classification algorithm
- artificial intelligence
- support vector machine svm
- partial occlusion
- global features
- test sample
- pose estimation
- support vector machine
- high resolution
- similarity measure
- computer vision
- watermarking technique
- nearest neighbor rule
- machine learning