File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1145/3381805
- Scopus: eid_2-s2.0-85089413000
- WOS: WOS:000582627100004
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: DEEPEYE: A Deeply Tensor-Compressed Neural Network for Video Comprehension on Terminal Devices
Title | DEEPEYE: A Deeply Tensor-Compressed Neural Network for Video Comprehension on Terminal Devices |
---|---|
Authors | |
Issue Date | 2020 |
Publisher | Association for Computing Machinery, Inc. |
Citation | ACM Transactions on Embedded Computing Systems, 2020, v. 19 n. 3, p. article no. 18 How to Cite? |
Abstract | Video object detection and action recognition typically require deep neural networks (DNNs) with huge number of parameters. It is thereby challenging to develop a DNN video comprehension unit in resource-constrained terminal devices. In this article, we introduce a deeply tensor-compressed video comprehension neural network, called DEEPEYE, for inference on terminal devices. Instead of building a Long Short-Term Memory (LSTM) network directly from high-dimensional raw video data input, we construct an LSTM-based spatio-temporal model from structured, tensorized time-series features for object detection and action recognition. A deep compression is achieved by tensor decomposition and trained quantization of the time-series feature-based LSTM network. We have implemented DEEPEYE on an ARM-core-based IOT board with 31 FPS consuming only 2.4W power. Using the video datasets MOMENTS, UCF11 and HMDB51 as benchmarks, DEEPEYE achieves a 228.1× model compression with only 0.47% mAP reduction; as well as 15k× parameter reduction with up to 8.01% accuracy improvement over other competing approaches. |
Persistent Identifier | http://hdl.handle.net/10722/289689 |
ISSN | 2023 Impact Factor: 2.8 2023 SCImago Journal Rankings: 0.830 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Cheng, Y | - |
dc.contributor.author | Li, G | - |
dc.contributor.author | Wong, N | - |
dc.contributor.author | Chen, HB | - |
dc.contributor.author | Yu, H | - |
dc.date.accessioned | 2020-10-22T08:16:03Z | - |
dc.date.available | 2020-10-22T08:16:03Z | - |
dc.date.issued | 2020 | - |
dc.identifier.citation | ACM Transactions on Embedded Computing Systems, 2020, v. 19 n. 3, p. article no. 18 | - |
dc.identifier.issn | 1539-9087 | - |
dc.identifier.uri | http://hdl.handle.net/10722/289689 | - |
dc.description.abstract | Video object detection and action recognition typically require deep neural networks (DNNs) with huge number of parameters. It is thereby challenging to develop a DNN video comprehension unit in resource-constrained terminal devices. In this article, we introduce a deeply tensor-compressed video comprehension neural network, called DEEPEYE, for inference on terminal devices. Instead of building a Long Short-Term Memory (LSTM) network directly from high-dimensional raw video data input, we construct an LSTM-based spatio-temporal model from structured, tensorized time-series features for object detection and action recognition. A deep compression is achieved by tensor decomposition and trained quantization of the time-series feature-based LSTM network. We have implemented DEEPEYE on an ARM-core-based IOT board with 31 FPS consuming only 2.4W power. Using the video datasets MOMENTS, UCF11 and HMDB51 as benchmarks, DEEPEYE achieves a 228.1× model compression with only 0.47% mAP reduction; as well as 15k× parameter reduction with up to 8.01% accuracy improvement over other competing approaches. | - |
dc.language | eng | - |
dc.publisher | Association for Computing Machinery, Inc. | - |
dc.relation.ispartof | ACM Transactions on Embedded Computing Systems | - |
dc.rights | ACM Transactions on Embedded Computing Systems. Copyright © Association for Computing Machinery, Inc. | - |
dc.rights | ©ACM, YYYY. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in PUBLICATION, {VOL#, ISS#, (DATE)} http://doi.acm.org/10.1145/nnnnnn.nnnnnn | - |
dc.title | DEEPEYE: A Deeply Tensor-Compressed Neural Network for Video Comprehension on Terminal Devices | - |
dc.type | Article | - |
dc.identifier.email | Cheng, Y: cyuan328@hku.hk | - |
dc.identifier.email | Wong, N: nwong@eee.hku.hk | - |
dc.identifier.authority | Wong, N=rp00190 | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1145/3381805 | - |
dc.identifier.scopus | eid_2-s2.0-85089413000 | - |
dc.identifier.hkuros | 315877 | - |
dc.identifier.volume | 19 | - |
dc.identifier.issue | 3 | - |
dc.identifier.spage | article no. 18 | - |
dc.identifier.epage | article no. 18 | - |
dc.identifier.isi | WOS:000582627100004 | - |
dc.publisher.place | United States | - |
dc.identifier.issnl | 1539-9087 | - |