File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/CVPR42600.2020.00158
- Scopus: eid_2-s2.0-85094857808
- WOS: WOS:000620679501075
- Find via
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: Multi-Dimensional Pruning: A Unified Framework for Model Compression
Title | Multi-Dimensional Pruning: A Unified Framework for Model Compression |
---|---|
Authors | |
Issue Date | 2020 |
Citation | Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2020, p. 1505-1514 How to Cite? |
Abstract | In this work, we propose a unified model compression framework called Multi-Dimensional Pruning (MDP) to simultaneously compress the convolutional neural networks (CNNs) on multiple dimensions. In contrast to the existing model compression methods that only aim to reduce the redundancy along either the spatial/spatial-temporal dimension (e.g., spatial dimension for 2D CNNs, spatial and temporal dimensions for 3D CNNs) or the channel dimension, our newly proposed approach can simultaneously reduce the spatial/spatial-temporal and the channel redundancies for CNNs. Specifically, in order to reduce the redundancy along the spatial/spatial-temporal dimension, we downsample the input tensor of a convolutional layer, in which the scaling factor for the downsampling operation is adaptively selected by our approach. After the convolution operation, the output tensor is upsampled to the original size to ensure the unchanged input size for the subsequent CNN layers. To reduce the channel-wise redundancy, we introduce a gate for each channel of the output tensor as its importance score, in which the gate value is automatically learned. The channels with small importance scores will be removed after the model compression process. Our comprehensive experiments on four benchmark datasets demonstrate that our MDP framework outperforms the existing methods when pruning both 2D CNNs and 3D CNNs. |
Persistent Identifier | http://hdl.handle.net/10722/321905 |
ISSN | 2020 SCImago Journal Rankings: 4.658 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Guo, Jinyang | - |
dc.contributor.author | Ouyang, Wanli | - |
dc.contributor.author | Xu, Dong | - |
dc.date.accessioned | 2022-11-03T02:22:15Z | - |
dc.date.available | 2022-11-03T02:22:15Z | - |
dc.date.issued | 2020 | - |
dc.identifier.citation | Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2020, p. 1505-1514 | - |
dc.identifier.issn | 1063-6919 | - |
dc.identifier.uri | http://hdl.handle.net/10722/321905 | - |
dc.description.abstract | In this work, we propose a unified model compression framework called Multi-Dimensional Pruning (MDP) to simultaneously compress the convolutional neural networks (CNNs) on multiple dimensions. In contrast to the existing model compression methods that only aim to reduce the redundancy along either the spatial/spatial-temporal dimension (e.g., spatial dimension for 2D CNNs, spatial and temporal dimensions for 3D CNNs) or the channel dimension, our newly proposed approach can simultaneously reduce the spatial/spatial-temporal and the channel redundancies for CNNs. Specifically, in order to reduce the redundancy along the spatial/spatial-temporal dimension, we downsample the input tensor of a convolutional layer, in which the scaling factor for the downsampling operation is adaptively selected by our approach. After the convolution operation, the output tensor is upsampled to the original size to ensure the unchanged input size for the subsequent CNN layers. To reduce the channel-wise redundancy, we introduce a gate for each channel of the output tensor as its importance score, in which the gate value is automatically learned. The channels with small importance scores will be removed after the model compression process. Our comprehensive experiments on four benchmark datasets demonstrate that our MDP framework outperforms the existing methods when pruning both 2D CNNs and 3D CNNs. | - |
dc.language | eng | - |
dc.relation.ispartof | Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | - |
dc.title | Multi-Dimensional Pruning: A Unified Framework for Model Compression | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/CVPR42600.2020.00158 | - |
dc.identifier.scopus | eid_2-s2.0-85094857808 | - |
dc.identifier.spage | 1505 | - |
dc.identifier.epage | 1514 | - |
dc.identifier.isi | WOS:000620679501075 | - |