Sign language recognition using sub-units
Citations
382 citations
Cites methods from "Sign language recognition using sub..."
...Until recently SLR methods have mainly used handcrafted intermediate representations [33, 16] and the temporal changes in these features have been modelled using classical graph based approaches, such as Hidden Markov Models (HMMs) [58], Conditional Random Fields [62] or template based methods [5, 48]....
[...]
333 citations
Cites methods from "Sign language recognition using sub..."
...The approach in [4] uses the Microsoft Kinect to extract appearance-based hand features and track the position in 2D and 3D....
[...]
309 citations
Additional excerpts
...[7] compare...
[...]
263 citations
255 citations
Cites background from "Sign language recognition using sub..."
...Continuous sign language recognition is different from isolated gesture classification [7, 20] or sign spotting [8, 23, 30], which is to detect predefined signs from video stream and the supervision contains exact temporal locations for each sign....
[...]
References
[...]
79,257 citations
"Sign language recognition using sub..." refers methods in this paper
...Random forests were proposed by Amit and Geman (1997) and Breiman (2001)....
[...]
18,620 citations
"Sign language recognition using sub..." refers methods in this paper
...Their later work (Vogler and Metaxas, 1999), used parallel HMMs on both hand shape and motion sub-units, similar to those proposed by the linguist Stokoe (1960)....
[...]
...The Viola Jones face detector (Viola and Jones, 2001) is used to locate the face....
[...]
15,813 citations
7,963 citations
"Sign language recognition using sub..." refers methods in this paper
...Finally, the Hu set of invariant moments are considered, there are 7 of these moments and they are created by combining the normalised central moments, see Hu (1962) for full details, they offer invariance to scale, translation, rotation and skew....
[...]
...Four types were chosen to form a feature vector, m: spatial, mab, central, μab, normalised central, μ̄ab and the Hu set of invariant moments (Hu, 1962) H1-H7. The order of a moment is defined as a+ b. This work uses all moments, central moments and normalised central moments up to the 3rd order, 10 per type, (00, 01, 10, 11, 20, 02, 12, 21, 30, 03). Finally, the Hu set of invariant moments are considered, there are 7 of these moments and they are created by combining the normalised central moments, see Hu (1962) for full details, they offer invariance to scale, translation, rotation and skew....
[...]
...Four types were chosen to form a feature vector, m: spatial, mab, central, µab, normalised central, µ̄ab and the Hu set of invariant moments (Hu, 1962) H1-H7....
[...]
...Four types were chosen to form a feature vector, m: spatial, mab, central, μab, normalised central, μ̄ab and the Hu set of invariant moments (Hu, 1962) H1-H7....
[...]
1,214 citations
"Sign language recognition using sub..." refers background or methods in this paper
...Random forests were proposed by Amit and Geman (1997) and Breiman (2001)....
[...]
...Random forests were proposed by Amit and Geman (1997) and Breiman (2001). They have been shown to yield good performance on a variety of classification and regression problems, and can be trained efficiently in a parallel manner, allowing training on large feature vectors and data sets....
[...]