Evaluating spatiotemporal interest point features for depth-based action recognition

By Y. Zhu, Wenbin Chen, Guo-qing Guo. 2014

121 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Human-Computer Interaction

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. German Ignacio Parisi
2. Stefan Wermter
3. Guo-qing Guo
4. Yingli Tian
5. Cornelius Weber
6. Weishi Zheng
7. J. Lai
8. Jianfang Hu

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Xiaogang Wang
3. Jun Liu
4. G. Wang
5. Dacheng Tao
6. Andrew Zisserman
7. Ming-Hsuan Yang
8. Yi Yang

In Human-Computer Interaction

1. Yifan Zhang
2. Hanqing Lu
3. Jian Cheng
4. Stefan Wermter
5. German Ignacio Parisi
6. Jun Wan
7. S. Li
8. Guo-qing Guo

Citations in the corpus

• RGB-D-based action recognition datasets: A survey (2016, neutral): “3. D 1. CS (Half training, half test) 2. CS (Half training, half test) 3. LOSubO CAD-60 [12] 1. Decision-Level Fusion (Majority Voting) [61] 2. Pose Kinectic Energy [70] 3. SpatioTemporal Interest Pt [71] 1. 96.4(Prec.) 84.6(Rec.) 2. 93.8(Prec.) 94.5(Rec.) 3. 93.2(Prec.) 84.6(Rec.) 1. D, S 2. S 3. D”
• Self-organizing neural integration of pose-motion features for human action recognition (2015, background): “Zhu et al. (2014) evaluated a set of spatio-temporal interest point features from raw depth map images to classify actions with a SVM. Experiments were conducted also using interest points in combination with skeleton joint positions and color information, obtaining better results.”
• 3D-based Deep Convolutional Neural Network for action recognition with depth sequences (2016, methodology): “Their experiments revealed that skeleton joint information can greatly improve performance by fusing spatio-temporal features and skeleton joints for depth-based action recognition [22].”
• 3D human action segmentation and recognition using pose kinetic energy (2014, unknown): “[42] 93.”
• EgoGesture: A New Dataset and Benchmark for Egocentric Hand Gesture Recognition (2018, background): “They can be extracted from the single frame or consecutive frame sequence locally at spatiotemporal interest points [27] or densely sampled in the whole frame [28].”
• Explore Efficient Local Features from RGB-D Data for One-Shot Learning Gesture Recognition (2016, methodology): “That is because: (1) we just used RGB-D data while other state-of-the-art methods [61], [62], [63] used skeleton information which is more easier to locate the human body.”
• Jointly Learning Heterogeneous Features for RGB-D Activity Recognition (2015, background): “ward way is to generalize the descriptors developed for RGB channel to depth channel, typically for describing the shape geometry [13, 28, 16, 41, 22, 36, 16], where Depth HOG [22, 36] is one representative.”
• A Human Activity Recognition System Using Skeleton Data from RGBD Sensors (2016, background): “This approach can be also extended to depth and skeleton data, combining the features with random forests [21].”
• Learning the Depths of Moving People by Watching Frozen People (2019, background): “Several action recognition methods use depth sensors to capture human actions [54, 34, 22, 26], but most use a static camera and provide only a limited number of indoor scenes.”
• 3D action recognition using data visualization and convolutional neural networks (2017, methodology): “Following [25], we use crosssubject validation with subjects 1,2,3,4,5 for training and subjects 6,7,8,9,10 for testing.”
• Hand Gesture Recognition in Real Time for Automotive Interfaces: A Multimodal Vision-Based Approach and Evaluations (2014, methodology): “extracted using spatiotemporal interest points (as in [22] and [23]) or densely sampled.”