Sparsity and persistence: mixed norms provide simple signal models with dependent coefficients

By Matthieu Kowalski, Bruno Torrésani. 2009

153 citations recorded by the source corpus.

Subfields: Civil and Structural Engineering, Computational Mechanics, Computer Vision and Pattern Recognition

How researchers see this paper

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

In Civil and Structural Engineering

1. Matthieu Kowalski
2. Shuicheng Yan
3. Xiao-Tong Yuan
4. Mohamed-Jalal Fadili
5. G. Peyr'e
6. Hugo Raguet
7. Bruno Torrésani

In Computational Mechanics

1. Shuicheng Yan
2. Xiao-Tong Yuan
3. Matthieu Kowalski
4. Bruno Torrésani
5. G. Peyr'e
6. Mohamed-Jalal Fadili
7. Hugo Raguet
8. Qingxiong Yang

In Computer Vision and Pattern Recognition

1. Amir Shahroudy
2. G. Wang
3. Shuicheng Yan
4. Jun Liu
5. Hossein Rahmani
6. Tian-Tsong Ng
7. Gang Wang
8. Xiao-Tong Yuan

Citations in the corpus

• Multimodal Multipart Learning for Action Recognition in Depth Videos (2015, background): “Combining these two regularization terms can be considered as a trade off between sparsity and persistence of features [42] based on their relations across the parts, modalities, and between the action classes.”
• Linguistic Structured Sparsity in Text Categorization (2014, background): “3Other structured regularizers include the fused lasso (Tibshirani et al., 2005) and the elitist lasso (Kowalski and Torresani, 2009). where λglas is a hyperparameter tuned on a development data, and λg is a group specific weight.”
• A Generalized Forward-Backward Splitting (2011, methodology): “A way to take into account those dependencies is to replace the absolute value of the coefficients in the (cid:96) 1 -norm by the (cid:96) 2 -norm of groups (or blocks ) of coefficients [64, 78, 50, 15, 39].”
• Visual classification with multi-task joint sparse representation (2010, methodology): “The application of APG algorithm to group/multi-task joint sparse learning has been addressed in [51]–[53].”
• Structured Sparsity: from Mixed Norms to Structured Shrinkage (2009, methodology): “Such a hierarchy can be exploited by a mixed norm penalty [7], [ 8 ], the functional to minimize taking the form:”
• Sparse regression using mixed norms (2009, methodology): “Relaxation (BCR) algorithm with mixed norms [ 18 ] provides a valuable alternative to the thresholded Landweber iteration presented here.”
• Semi-Supervised Learning with Spectral Graph Wavelets (2011, unknown): “For example, [3, 5] consider: S(f) := ∑ (u,v)∈E wu,v [f(v)− f(u)]2 = fTLf. (1)”