3D SMoSIFT: three-dimensional sparse motion scale invariant feature transform for activity recognition from RGB-D videos

By Jun Wan, Q. Ruan, Gaoyun An, Ruizhen Zhao. 2014

49 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Endocrinology, Diabetes and Metabolism

How researchers see this paper

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

In Computer Vision and Pattern Recognition

1. Pichao Wang
2. Wanqing Li
3. Philip O. Ogunbona
4. Zhimin Gao
5. Chang-Fu Tang
6. Jun Wan
7. Yuyao Zhang
8. Jun Liu

In Endocrinology, Diabetes and Metabolism

1. Pichao Wang
2. Wanqing Li
3. Philip O. Ogunbona
4. Zhimin Gao
5. Chang-Fu Tang
6. Jun Wan
7. Guo-qing Guo
8. Yuyao Zhang

Citations in the corpus

• Scene Flow to Action Map: A New Representation for RGB-D Based Action Recognition with Convolutional Neural Networks (2017, methodology): “Among these methods, MFSK combined 3D SMoSIFT [62] with (HOG, HOF and MBH) [64] descriptors.”
• Multimodal Gesture Recognition Using 3-D Convolution and Convolutional LSTM (2017, methodology): “[30] first proposed the 3-D enhancedmotion scale invariant feature transform (3D EMoSIFT) and 3-D Sparse Motion SIFT (3-D SMoSIFT) to extract spatiotemporal features from RGB-D images, and then the 3-D EMoSIFT and 3-D SMoSIFT features are evaluated under the bag of visual words (BoVW) model.”
• Multi-Modality Fusion based on Consensus-Voting and 3D Convolution for Isolated Gesture Recognition (2016, background): “For example, Wan et al [34] extracted a novel spatiotemporal feature named MFSK while [32] proposed to calculate SIFT-like descriptors on 3D gradient and motion spaces respectively for RGB-D video recognition.”
• Large-scale gesture recognition with a fusion of RGB-D data based on the C3D model (2016, background): “extend the scale invariant feature transform (SIFT) and propose 3D enhanced motion SIFT (3D EMoSIFT) [11] and 3D Sparse Motion SIFT (3D SMoSIFT) [12] to extract features with the fusion of RGBD data.”
• Multi-layered gesture recognition with Kinect (2015, methodology): “Wan et al. (2014a) proposes a discriminative dictionary learning method on 3D EMoSIFT features based on mutual information and then uses sparse reconstruction for classification.”
• Depth Pooling Based Large-Scale 3-D Action Recognition With Convolutional Neural Networks (2018, methodology): “These methods include MFSK combined 3D SMoSIFT [52] with (HOG, HOF and MBH) [53] descriptors.”
• Explore Efficient Local Features from RGB-D Data for One-Shot Learning Gesture Recognition (2016, methodology): “It shows some samples derived from [20] for the new subsets on CGD.”