Action recognition in still images using a multi-attention guided network with weakly supervised saliency detection.
Seyed Sajad AshrafiShahriar B. ShokouhiAhmad AyatollahiPublished in: Multim. Tools Appl. (2021)
Keyphrases
- action recognition
- object detectors
- image features
- three dimensional
- input image
- bag of words
- saliency map
- human actions
- image database
- image classification
- image data
- computer vision
- image retrieval
- multiple images
- object recognition
- test images
- lighting conditions
- bounding box
- image annotation
- human motion
- eye tracking
- multiscale
- viewpoint