Stable Low-Rank Tensor Decomposition for Compression of Convolutional Neural Network.
Anh Huy PhanKonstantin SobolevKonstantin SozykinDmitry ErmilovJulia GusakPetr TichavskýValeriy GlukhovIvan V. OseledetsAndrzej CichockiPublished in: ECCV (29) (2020)
Keyphrases
- tensor decomposition
- low rank
- convolutional neural network
- missing data
- convex optimization
- face detection
- linear combination
- matrix factorization
- high order
- matrix completion
- singular value decomposition
- high dimensional data
- low rank matrix
- semi supervised
- auxiliary information
- data representation
- rank minimization
- singular values
- data matrix
- neural network
- structure from motion
- visual data
- image data
- optical flow
- training data