IIRNet: A lightweight deep neural network using intensely inverted residuals for image recognition
Deep neural networks have achieved great success in many tasks of pattern recognition. However, large model size and high cost in computation limit their applications in resource-limited systems. In this paper, our focus is to design a lightweight and efficient convolutional neural network architect...
Uloženo v:
| Vydáno v: | Image and vision computing Ročník 92; s. 103819 |
|---|---|
| Hlavní autoři: | , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
Elsevier B.V
01.12.2019
|
| Témata: | |
| ISSN: | 0262-8856, 1872-8138 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Deep neural networks have achieved great success in many tasks of pattern recognition. However, large model size and high cost in computation limit their applications in resource-limited systems. In this paper, our focus is to design a lightweight and efficient convolutional neural network architecture by directly training the compact network for image recognition. To achieve a good balance among classification accuracy, model size, and computation complexity, we propose a lightweight convolutional neural network architecture named IIRNet for resource-limited systems. The new architecture is built based on Intensely Inverted Residual block (IIR block) to decrease the redundancy of the convolutional blocks. By utilizing two new operations, intensely inverted residual and multi-scale low-redundancy convolutions, IIR block greatly reduces its model size and computational costs while matches the classification accuracy of the state-of-the-art networks. Experiments on CIFAR-10, CIFAR-100, and ImageNet datasets demonstrate the superior performance of IIRNet on the trade-offs among classification accuracy, computation complexity, and model size, compared to the mainstream compact network architectures.
•A lightweight and efficient convolutional neural network architecture is constructed.•Intensely inverted residual and multi-scale low-redundancy convolutions are used to reduce the model size and complexity.•The proposed network achieves comparable classification accuracy to the mainstream compact network architectures.•Balanced performance is obtained on three challenging datasets. |
|---|---|
| AbstractList | Deep neural networks have achieved great success in many tasks of pattern recognition. However, large model size and high cost in computation limit their applications in resource-limited systems. In this paper, our focus is to design a lightweight and efficient convolutional neural network architecture by directly training the compact network for image recognition. To achieve a good balance among classification accuracy, model size, and computation complexity, we propose a lightweight convolutional neural network architecture named IIRNet for resource-limited systems. The new architecture is built based on Intensely Inverted Residual block (IIR block) to decrease the redundancy of the convolutional blocks. By utilizing two new operations, intensely inverted residual and multi-scale low-redundancy convolutions, IIR block greatly reduces its model size and computational costs while matches the classification accuracy of the state-of-the-art networks. Experiments on CIFAR-10, CIFAR-100, and ImageNet datasets demonstrate the superior performance of IIRNet on the trade-offs among classification accuracy, computation complexity, and model size, compared to the mainstream compact network architectures.
•A lightweight and efficient convolutional neural network architecture is constructed.•Intensely inverted residual and multi-scale low-redundancy convolutions are used to reduce the model size and complexity.•The proposed network achieves comparable classification accuracy to the mainstream compact network architectures.•Balanced performance is obtained on three challenging datasets. |
| ArticleNumber | 103819 |
| Author | Li, Yuyuan Zhang, Dong Lee, Dah-Jye |
| Author_xml | – sequence: 1 givenname: Yuyuan surname: Li fullname: Li, Yuyuan organization: School of Electronics and Information Technology, Sun Yat-sen University, Guangzhou, China – sequence: 2 givenname: Dong surname: Zhang fullname: Zhang, Dong email: zhangd@mail.sysu.edu.cn organization: School of Electronics and Information Technology, Sun Yat-sen University, Guangzhou, China – sequence: 3 givenname: Dah-Jye surname: Lee fullname: Lee, Dah-Jye organization: Sun Yat-sen University, Shunde Research Institute, Shunde, Foshan, China |
| BookMark | eNqFkM9OwzAMxiM0JMbgDTjkBVqSdkmzHZCmiT-TJpAQnKMkdUtGSack27S3J9U4cYCLbX32Z9m_SzRyvQOEbijJKaH8dpPbL7W3IS8InSUpJ4SdoTEVVZEJWooRGpOCp1owfoEuQ9gQQipSzcZIr1avzxDneIE7237EAwwR1wBb7GDnVZdSPPT-E--CdS22LoIL0B1TtQcfocYegq13qgu46T1Op7SQNNO3zkbbuyt03qQmXP_kCXp_uH9bPmXrl8fVcrHOTEl4zBqt-FTPhJ4WNW-YJsQYbagpWaOFUFBVNQOmeUOFZlwZoQ0TROmal4UpQJQTND_tNb4PwUMjjY1quCB6ZTtJiRxoyY080ZIDrUFNtJJ5-su89WnMH_-z3Z1skB7bW_AyGAvOQG0TgSjr3v694BtDsovr |
| CitedBy_id | crossref_primary_10_1109_JIOT_2025_3586661 crossref_primary_10_1007_s10489_023_04544_1 crossref_primary_10_1109_ACCESS_2023_3272985 crossref_primary_10_1109_JIOT_2021_3063497 crossref_primary_10_1007_s11282_023_00677_8 crossref_primary_10_1016_j_ress_2024_110145 crossref_primary_10_1038_s41598_025_93718_7 crossref_primary_10_3390_jne2040036 crossref_primary_10_1155_2021_4529107 crossref_primary_10_3390_math9111245 crossref_primary_10_1109_TCDS_2023_3311171 crossref_primary_10_1016_j_imavis_2024_105037 crossref_primary_10_1016_j_imavis_2025_105646 crossref_primary_10_1109_ACCESS_2022_3151660 crossref_primary_10_1007_s11760_025_03945_8 crossref_primary_10_1016_j_imavis_2021_104286 |
| Cites_doi | 10.1016/j.neucom.2018.10.075 10.1016/j.neucom.2018.10.070 10.1109/TPAMI.2008.128 10.1016/j.neucom.2018.11.028 10.1016/j.neucom.2018.11.010 10.1016/j.neucom.2018.11.031 10.1016/j.neucom.2018.10.071 10.1109/TPAMI.2015.2502579 |
| ContentType | Journal Article |
| Copyright | 2019 Elsevier B.V. |
| Copyright_xml | – notice: 2019 Elsevier B.V. |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.imavis.2019.10.005 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Applied Sciences Engineering |
| EISSN | 1872-8138 |
| ExternalDocumentID | 10_1016_j_imavis_2019_10_005 S0262885619301441 |
| GroupedDBID | --K --M .~1 0R~ 1B1 1~. 1~5 29I 4.4 457 4G. 5GY 5VS 7-5 71M 8P~ 9JN AABNK AACTN AAEDT AAEDW AAIAV AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AAXUO AAYFN ABBOA ABFNM ABFRF ABJNI ABMAC ABOCM ABTAH ABXDB ABYKQ ACDAQ ACGFO ACGFS ACNNM ACRLP ACZNC ADBBV ADEZE ADJOM ADMUD ADTZH AEBSH AECPX AEFWE AEKER AENEX AFKWA AFTJW AGHFR AGUBO AGYEJ AHHHB AHJVU AHZHX AIALX AIEXJ AIKHN AITUG AJBFU AJOXV ALMA_UNASSIGNED_HOLDINGS AMFUW AMRAJ AOUOD ASPBG AVWKF AXJTR AZFZN BJAXD BKOJK BLXMC CS3 DU5 EBS EFJIC EFLBG EJD EO8 EO9 EP2 EP3 F0J F5P FDB FEDTE FGOYB FIRID FNPLU FYGXN G-Q G8K GBLVA GBOLZ HLZ HVGLF HZ~ IHE J1W JJJVA KOM LG9 M41 MO0 N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 R2- RIG RNS ROL RPZ SBC SDF SDG SDP SES SEW SPC SPCBC SST SSV SSZ T5K TN5 UHS UNMZH VOH WUQ XFK XPP ZMT ZY4 ~G- 9DU AATTM AAXKI AAYWO AAYXX ABDPE ABWVN ACLOT ACRPL ACVFH ADCNI ADNMO AEIPS AEUPX AFJKZ AFPUW AGQPQ AIGII AIIUN AKBMS AKRWK AKYEP ANKPU APXCP CITATION EFKBS ~HD |
| ID | FETCH-LOGICAL-c306t-fba64b98b42d6f5b00ccbc1c35fb88ae77d5e5b6f18b56ac8bc580abd632c2e83 |
| ISICitedReferencesCount | 20 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000502884200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0262-8856 |
| IngestDate | Tue Nov 18 22:13:35 EST 2025 Sat Nov 29 07:19:19 EST 2025 Fri Feb 23 02:48:36 EST 2024 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Lightweight CNN Computation complexity Image recognition Model size Low-redundancy Convolutional neural network (CNN) |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c306t-fba64b98b42d6f5b00ccbc1c35fb88ae77d5e5b6f18b56ac8bc580abd632c2e83 |
| ParticipantIDs | crossref_citationtrail_10_1016_j_imavis_2019_10_005 crossref_primary_10_1016_j_imavis_2019_10_005 elsevier_sciencedirect_doi_10_1016_j_imavis_2019_10_005 |
| PublicationCentury | 2000 |
| PublicationDate | December 2019 2019-12-00 |
| PublicationDateYYYYMMDD | 2019-12-01 |
| PublicationDate_xml | – month: 12 year: 2019 text: December 2019 |
| PublicationDecade | 2010 |
| PublicationTitle | Image and vision computing |
| PublicationYear | 2019 |
| Publisher | Elsevier B.V |
| Publisher_xml | – name: Elsevier B.V |
| References | Lin, Chen, Yan (bb0195) 2014 Torralba, Fergus, Freeman (bb0180) 2008; 30 He, Zhang, Ren, Sun (bb0210) 2016 Gholami, Kwon, Wu (bb0255) 2018 He, Zhang, Ren, Sun (bb0040) 2016 Wang, Wei, Zhang, Zeng (bb0220) 2016 Yang, Huang, Lv (bb0015) 2019; 330 Chollet (bb0160) 2017 Novikov, Podoprikhin, Osokin, Vetrov (bb0060) 2015 Lebedev, Ganin, Rakhuba, Oseledets, Lempitsky (bb0050) 2015 Huang, Li, Pleiss, Liu, Hopcroft, Weinberger (bb0230) 2017 Chen, Choi, Yu, Han, Chandraker (bb0110) 2017 Li, Zhang, Lee (bb0130) 2019; 329 Krizhevsky, Hinton (bb0175) 2009 Srivastava, Hinton, Krizhevsky, Sutskever, Salakhutdinov (bb0170) 2014; 15 Singh, Hoiem, Forsyth (bb0250) 2016 Ren, Pokrovsky, Yang, Urtasun (bb0070) 2018 Romero, Ballas, Kahou, Chassang, Gatta, Bengio (bb0200) 2015 Zagoruyko, Komodakis (bb0235) 2016 Yim, Joo, Bae, Kim (bb0105) 2017 Krizhevsky, Sutskever, Hinton (bb0025) 2012; 60 Yoon, Hwang (bb0075) 2017; 70 Luo, Wu, Lin (bb0080) 2017 He, Zhang, Sun (bb0085) 2017 Targ, Almeida, Lyman (bb0240) 2016 Zhang, Zou, He, Sun (bb0045) 2016; 38 Xie, Wang, Zhang, Lai, Hong, Qi (bb0140) 2018 Srivastava, Greff, Schmidhuber (bb0205) 2015 Lee, Xie, Gallagher, Zhang, Tu (bb0185) 2014 Wang, Sun, Eriksson, Wang (bb0065) 2018 Taheri, Toygar (bb0005) 2019; 329 Leng, Dou, Li, Zhu, Jin (bb0090) 2018 Yuan, Guo, Feng, Zhao, Xu, Wang, Choe, Duan (bb0020) 2019; 330 Howard, Zhu, Chen, Kalenichenko, Wang, Weyand, Andreetto, Adam (bb0115) 2017 Iandola, Han, Moskewicz, Ashraf, Dally, Keutzer (bb0150) 2016 Sun, Li, Liu, Wang (bb0155) 2018 Wang, Hu, Zhang, Zhang, Liu, Cheng (bb0095) 2018 Huang, Sun, Liu, Sedra, Weinberger (bb0215) 2016 Szegedy, Liu, Jia, Sermanet, Reed, Anguelov, Erhan, Vanhoucke, Rabinovich (bb0035) 2015 Sandler, Howard, Zhu, Zhmoginov, Chen (bb0120) 2018 Zhang, Zhou, Lin, Sun (bb0125) 2018 Szegedy, Vanhoucke, Ioffe, Shlens, Wojna (bb0165) 2016 Ou, Li (bb0145) 2019; 330 Zhang, Qi, Xiao, Wang (bb0135) 2017 Jia, Shelhamer, Donahue, Karayev, Long, Girshick, Guadarrama, Darrell (bb0190) 2014 Alistarh, Grubic, Li, Tomioka, Vojnovic (bb0100) 2017 Larsson, Maire, Shakhnarovich (bb0225) 2016 Zhao, Wang, Li, Tu (bb0245) 2016 Kim, Park, Yoo, Choi, Yang, Shin (bb0055) 2016 Simonyan, Zisserman (bb0030) 2015 Brahimi, Aoun, Amar (bb0010) 2019; 330 Wang (10.1016/j.imavis.2019.10.005_bb0095) 2018 Wang (10.1016/j.imavis.2019.10.005_bb0065) 2018 Yoon (10.1016/j.imavis.2019.10.005_bb0075) 2017; 70 Li (10.1016/j.imavis.2019.10.005_bb0130) 2019; 329 Torralba (10.1016/j.imavis.2019.10.005_bb0180) 2008; 30 Chollet (10.1016/j.imavis.2019.10.005_bb0160) 2017 Yang (10.1016/j.imavis.2019.10.005_bb0015) 2019; 330 He (10.1016/j.imavis.2019.10.005_bb0040) 2016 Zhang (10.1016/j.imavis.2019.10.005_bb0045) 2016; 38 Szegedy (10.1016/j.imavis.2019.10.005_bb0165) 2016 He (10.1016/j.imavis.2019.10.005_bb0210) 2016 Srivastava (10.1016/j.imavis.2019.10.005_bb0205) 2015 Zhang (10.1016/j.imavis.2019.10.005_bb0135) 2017 Yuan (10.1016/j.imavis.2019.10.005_bb0020) 2019; 330 Brahimi (10.1016/j.imavis.2019.10.005_bb0010) 2019; 330 Luo (10.1016/j.imavis.2019.10.005_bb0080) 2017 Singh (10.1016/j.imavis.2019.10.005_bb0250) 2016 Targ (10.1016/j.imavis.2019.10.005_bb0240) 2016 Chen (10.1016/j.imavis.2019.10.005_bb0110) 2017 Krizhevsky (10.1016/j.imavis.2019.10.005_bb0025) 2012; 60 Sun (10.1016/j.imavis.2019.10.005_bb0155) 2018 Huang (10.1016/j.imavis.2019.10.005_bb0230) 2017 Xie (10.1016/j.imavis.2019.10.005_bb0140) 2018 Iandola (10.1016/j.imavis.2019.10.005_bb0150) 2016 Novikov (10.1016/j.imavis.2019.10.005_bb0060) 2015 Jia (10.1016/j.imavis.2019.10.005_bb0190) 2014 Wang (10.1016/j.imavis.2019.10.005_bb0220) 2016 He (10.1016/j.imavis.2019.10.005_bb0085) 2017 Srivastava (10.1016/j.imavis.2019.10.005_bb0170) 2014; 15 Howard (10.1016/j.imavis.2019.10.005_bb0115) 2017 Gholami (10.1016/j.imavis.2019.10.005_bb0255) 2018 Taheri (10.1016/j.imavis.2019.10.005_bb0005) 2019; 329 Alistarh (10.1016/j.imavis.2019.10.005_bb0100) 2017 Yim (10.1016/j.imavis.2019.10.005_bb0105) 2017 Kim (10.1016/j.imavis.2019.10.005_bb0055) 2016 Leng (10.1016/j.imavis.2019.10.005_bb0090) 2018 Szegedy (10.1016/j.imavis.2019.10.005_bb0035) 2015 Zagoruyko (10.1016/j.imavis.2019.10.005_bb0235) 2016 Simonyan (10.1016/j.imavis.2019.10.005_bb0030) 2015 Lee (10.1016/j.imavis.2019.10.005_bb0185) 2014 Sandler (10.1016/j.imavis.2019.10.005_bb0120) 2018 Romero (10.1016/j.imavis.2019.10.005_bb0200) 2015 Lebedev (10.1016/j.imavis.2019.10.005_bb0050) 2015 Ou (10.1016/j.imavis.2019.10.005_bb0145) 2019; 330 Zhao (10.1016/j.imavis.2019.10.005_bb0245) 2016 Zhang (10.1016/j.imavis.2019.10.005_bb0125) 2018 Lin (10.1016/j.imavis.2019.10.005_bb0195) 2014 Larsson (10.1016/j.imavis.2019.10.005_bb0225) 2016 Krizhevsky (10.1016/j.imavis.2019.10.005_bb0175) 2009 Ren (10.1016/j.imavis.2019.10.005_bb0070) 2018 Huang (10.1016/j.imavis.2019.10.005_bb0215) 2016 |
| References_xml | – start-page: 1707 year: 2017 end-page: 1718 ident: bb0100 article-title: QSGD: communication-efficient SGD via gradient quantization and encoding publication-title: Advances in Neural Information Processing Systems – volume: 38 start-page: 1943 year: 2016 end-page: 1955 ident: bb0045 article-title: Accelerating very deep convolutional networks for classification and detection publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – start-page: 770 year: 2016 end-page: 778 ident: bb0040 article-title: Deep residual learning for image recognition publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – start-page: 28 year: 2016 end-page: 36 ident: bb0250 article-title: Swapout: learning an ensemble of deep architectures publication-title: Advances in Neural Information Processing Systems – volume: 330 start-page: 253 year: 2019 end-page: 258 ident: bb0145 article-title: Vector-kernel convolutional neural networks publication-title: Neurocomputing – start-page: 646 year: 2016 end-page: 661 ident: bb0215 article-title: Deep networks with stochastic depth publication-title: Proceedings of the European Conference on Computer Vision – volume: 60 start-page: 1097 year: 2012 end-page: 1105 ident: bb0025 article-title: ImageNet classification with deep convolutional neural networks publication-title: Advances in Neural Information Processing Systems – start-page: 442 year: 2015 end-page: 450 ident: bb0060 article-title: Tensorizing neural networks publication-title: Proceedings of the Neural Information Processing Systems – start-page: 742 year: 2017 end-page: 751 ident: bb0110 article-title: Learning efficient object detection models with knowledge distillation publication-title: Advances in Neural Information Processing Systems – year: 2017 ident: bb0230 article-title: Snapshot Ensembles: Train 1, Get m for Free – start-page: 1 year: 2016 end-page: 4 ident: bb0240 article-title: Resnet in resnet: generalizing residual architectures publication-title: Proceedings of the International Conference on Learning Representations – year: 2018 ident: bb0140 article-title: IGCV2: interleaved structured sparse convolutional neural networks publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – start-page: 8711 year: 2018 end-page: 8720 ident: bb0070 article-title: SBNet: Sparse Blocks Network for fast inference publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2018 ident: bb0065 article-title: Wide compression: tensor ring nets publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2015 ident: bb0050 article-title: Speeding-up convolutional neural networks using fine-tuned CP-decomposition publication-title: Proceedings of the International Conference on Learning Representations – year: 2016 ident: bb0225 article-title: Fractalnet: Ultra-deep Neural Networks Without Residuals – volume: 15 start-page: 1929 year: 2014 end-page: 1958 ident: bb0170 article-title: Dropout: a simple way to prevent neural networks from overfitting publication-title: J. Mach. Learn. Res. – start-page: 1 year: 2015 end-page: 9 ident: bb0035 article-title: Going deeper with convolutions publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2018 ident: bb0095 article-title: Two-step quantization for low-bit neural networks publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2017 ident: bb0160 article-title: Xception: deep learning with depthwise separable convolutions publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2018 ident: bb0090 article-title: Extremely low bit neural network: squeeze the last bit out with ADMM publication-title: Proceedings of the Thirty-second AAAI Conference on Artificial Intelligence – year: 2017 ident: bb0135 article-title: Interleaved group convolutions for deep neural networks publication-title: Proceedings of the IEEE International Conference on Computer Vision – volume: 330 start-page: 127 year: 2019 end-page: 137 ident: bb0020 article-title: A jointly learned deep embedding for person re-identification publication-title: Neurocomputing – year: 2017 ident: bb0115 article-title: MobileNets: Efficient Convolutional Neural Networks for Mobile Vision – year: 2014 ident: bb0190 article-title: Caffe: Convolutional Architecture for Fast Feature Embedding – volume: 70 start-page: 3958 year: 2017 end-page: 3966 ident: bb0075 article-title: Combined group and exclusive sparsity for deep neural networks publication-title: Proceedings of the 34th International Conference on Machine Learning – volume: 30 start-page: 1958 year: 2008 end-page: 1970 ident: bb0180 article-title: 80 million tiny images: a large data set for nonparametric object and scene recognition publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – start-page: 5068 year: 2017 end-page: 5076 ident: bb0080 article-title: ThiNet: a filter level pruning method for deep neural network compression publication-title: Proceedings of the IEEE International Conference on Computer Vision – start-page: 4510 year: 2018 end-page: 4520 ident: bb0120 article-title: MobileNetV2: inverted residuals and linear bottlenecks publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2016 ident: bb0150 article-title: SqueezeNet: AlexNet-level Accuracy With 50X Fewer Parameters and < – volume: 329 start-page: 329 year: 2019 end-page: 338 ident: bb0130 article-title: Automatic fabric defect detection with a wide-and-compact network publication-title: Neurocomputing – start-page: 2818 year: 2016 end-page: 2826 ident: bb0165 article-title: Rethinking the inception architecture for computer vision publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – volume: 330 start-page: 48 year: 2019 end-page: 55 ident: bb0015 article-title: A feature learning approach for face recognition with robustness to noisy label based on top-N prediction publication-title: Neurocomputing – year: 2014 ident: bb0185 publication-title: Deeply-supervised Nets – year: 2016 ident: bb0235 article-title: Wide Residual Networks – year: 2015 ident: bb0205 publication-title: Highway Networks – year: 2015 ident: bb0030 article-title: Very deep convolutional networks for large-scale image recognition publication-title: Proceedings of the International Conference on Learning Representations – start-page: 630 year: 2016 end-page: 645 ident: bb0210 article-title: Identity mappings in deep residual networks publication-title: Proceedings of the European Conference on Computer Vision – year: 2018 ident: bb0125 article-title: ShuffleNet: an extremely efficient convolutional neural network for mobile devices publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – volume: 330 start-page: 337 year: 2019 end-page: 354 ident: bb0010 article-title: Boosted convolutional neural network for object recognition at large scale publication-title: Neurocomputing – volume: 329 start-page: 300 year: 2019 end-page: 310 ident: bb0005 article-title: On the use of DAG-CNN architecture for age estimation with multi-stage features fusion publication-title: Neurocomputing – start-page: 1 year: 2009 end-page: 60 ident: bb0175 article-title: Learning Multiple Layers of Features From Tiny Images – year: 2018 ident: bb0255 article-title: SqueezeNext: Hardware-aware Neural Network Design – year: 2016 ident: bb0055 article-title: Compression of deep convolutional neural networks for fast and low power mobile applications publication-title: Proceedings of the International Conference on Learning Representations – year: 2018 ident: bb0155 article-title: IGCV3: interleaved low-rank group convolutions for efficient deep neural networks publication-title: Proceedings of the 29 – start-page: 1389 year: 2017 end-page: 1397 ident: bb0085 article-title: Channel pruning for accelerating very deep neural networks publication-title: Proceedings of the IEEE International Conference on Computer Vision – start-page: 7130 year: 2017 end-page: 7138 ident: bb0105 article-title: A gift from knowledge distillation: fast optimization, network minimization and transfer learning publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – year: 2015 ident: bb0200 article-title: FITNETS: hints for thin deep nets publication-title: Proceedings of the International Conference on Learning Representations – year: 2016 ident: bb0245 article-title: Deep Convolutional Neural Networks With Merge-and-run Mappings – year: 2014 ident: bb0195 article-title: Network in network publication-title: Proceedings of the International Conference on Learning Representations – year: 2016 ident: bb0220 publication-title: Deeply-fused Nets – start-page: 770 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0040 article-title: Deep residual learning for image recognition – start-page: 742 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0110 article-title: Learning efficient object detection models with knowledge distillation – year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0230 – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0220 publication-title: Deeply-fused Nets – start-page: 1389 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0085 article-title: Channel pruning for accelerating very deep neural networks – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0150 – start-page: 1707 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0100 article-title: QSGD: communication-efficient SGD via gradient quantization and encoding – start-page: 7130 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0105 article-title: A gift from knowledge distillation: fast optimization, network minimization and transfer learning – volume: 330 start-page: 48 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0015 article-title: A feature learning approach for face recognition with robustness to noisy label based on top-N prediction publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.10.075 – volume: 329 start-page: 329 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0130 article-title: Automatic fabric defect detection with a wide-and-compact network publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.10.070 – volume: 30 start-page: 1958 issue: 11 year: 2008 ident: 10.1016/j.imavis.2019.10.005_bb0180 article-title: 80 million tiny images: a large data set for nonparametric object and scene recognition publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2008.128 – volume: 60 start-page: 1097 issue: 2 year: 2012 ident: 10.1016/j.imavis.2019.10.005_bb0025 article-title: ImageNet classification with deep convolutional neural networks publication-title: Advances in Neural Information Processing Systems – year: 2014 ident: 10.1016/j.imavis.2019.10.005_bb0190 – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0155 article-title: IGCV3: interleaved low-rank group convolutions for efficient deep neural networks – start-page: 8711 year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0070 article-title: SBNet: Sparse Blocks Network for fast inference – year: 2014 ident: 10.1016/j.imavis.2019.10.005_bb0185 publication-title: Deeply-supervised Nets – volume: 330 start-page: 253 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0145 article-title: Vector-kernel convolutional neural networks publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.11.028 – start-page: 1 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0240 article-title: Resnet in resnet: generalizing residual architectures – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0090 article-title: Extremely low bit neural network: squeeze the last bit out with ADMM – start-page: 2818 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0165 article-title: Rethinking the inception architecture for computer vision – start-page: 442 year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0060 article-title: Tensorizing neural networks – volume: 330 start-page: 127 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0020 article-title: A jointly learned deep embedding for person re-identification publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.11.010 – volume: 15 start-page: 1929 issue: 1 year: 2014 ident: 10.1016/j.imavis.2019.10.005_bb0170 article-title: Dropout: a simple way to prevent neural networks from overfitting publication-title: J. Mach. Learn. Res. – year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0205 publication-title: Highway Networks – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0235 – volume: 70 start-page: 3958 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0075 article-title: Combined group and exclusive sparsity for deep neural networks – year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0030 article-title: Very deep convolutional networks for large-scale image recognition – year: 2014 ident: 10.1016/j.imavis.2019.10.005_bb0195 article-title: Network in network – start-page: 646 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0215 article-title: Deep networks with stochastic depth – year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0200 article-title: FITNETS: hints for thin deep nets – year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0050 article-title: Speeding-up convolutional neural networks using fine-tuned CP-decomposition – start-page: 28 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0250 article-title: Swapout: learning an ensemble of deep architectures – start-page: 4510 year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0120 article-title: MobileNetV2: inverted residuals and linear bottlenecks – volume: 330 start-page: 337 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0010 article-title: Boosted convolutional neural network for object recognition at large scale publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.11.031 – start-page: 5068 year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0080 article-title: ThiNet: a filter level pruning method for deep neural network compression – volume: 329 start-page: 300 year: 2019 ident: 10.1016/j.imavis.2019.10.005_bb0005 article-title: On the use of DAG-CNN architecture for age estimation with multi-stage features fusion publication-title: Neurocomputing doi: 10.1016/j.neucom.2018.10.071 – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0255 – start-page: 1 year: 2015 ident: 10.1016/j.imavis.2019.10.005_bb0035 article-title: Going deeper with convolutions – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0065 article-title: Wide compression: tensor ring nets – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0140 article-title: IGCV2: interleaved structured sparse convolutional neural networks – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0055 article-title: Compression of deep convolutional neural networks for fast and low power mobile applications – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0095 article-title: Two-step quantization for low-bit neural networks – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0245 – year: 2018 ident: 10.1016/j.imavis.2019.10.005_bb0125 article-title: ShuffleNet: an extremely efficient convolutional neural network for mobile devices – start-page: 630 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0210 article-title: Identity mappings in deep residual networks – year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0160 article-title: Xception: deep learning with depthwise separable convolutions – start-page: 1 year: 2009 ident: 10.1016/j.imavis.2019.10.005_bb0175 – year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0135 article-title: Interleaved group convolutions for deep neural networks – volume: 38 start-page: 1943 issue: 10 year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0045 article-title: Accelerating very deep convolutional networks for classification and detection publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2015.2502579 – year: 2017 ident: 10.1016/j.imavis.2019.10.005_bb0115 – year: 2016 ident: 10.1016/j.imavis.2019.10.005_bb0225 |
| SSID | ssj0007079 |
| Score | 2.3960495 |
| Snippet | Deep neural networks have achieved great success in many tasks of pattern recognition. However, large model size and high cost in computation limit their... |
| SourceID | crossref elsevier |
| SourceType | Enrichment Source Index Database Publisher |
| StartPage | 103819 |
| SubjectTerms | Computation complexity Convolutional neural network (CNN) Image recognition Lightweight CNN Low-redundancy Model size |
| Title | IIRNet: A lightweight deep neural network using intensely inverted residuals for image recognition |
| URI | https://dx.doi.org/10.1016/j.imavis.2019.10.005 |
| Volume | 92 |
| WOSCitedRecordID | wos000502884200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVESC databaseName: Elsevier SD Freedom Collection Journals 2021 customDbUrl: eissn: 1872-8138 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0007079 issn: 0262-8856 databaseCode: AIEXJ dateStart: 19950201 isFulltext: true titleUrlDefault: https://www.sciencedirect.com providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9NAEF6hlAMceBQQ5aU9cLO2ih_rXXOLaBGpUIRQQeFk7VO0Cm5Ux9D-e2YfdiyKCj1wsaKNvV57vszOTL6ZQeh16jrR0EIRpUVOClopImyRkULkFrxmo7jP4v_ygS0WfLmsPsaAfuvbCbCm4RcX1fq_ihrGQNgudfYG4h4mhQH4DEKHI4gdjv8k-Pn808KEgF-ycq73Tx_9TLQx68SVrwShNIH8nXRtSGlxNHazcimArj2zcRktrc_S8tUakpPvjtkzcI2iJKNJO_dfuvh7yFP3LPVu02-JjuzjGQNfu8tuC8UhUH1wNjoxsIIOxDdydGnGAYm0GpE7gt7KSlCyPNQL75VsaHgXtaQryh4U5RUFHmIJp_vwYLBoR72r9j37jm43rP5P-t_2sYFd2BPXTuswS-1mgaHaF7vdyRit-ATtzOaHy6Nh13aVAkM8Liy-T7P0XMCrq_mzGTMyTY4foHvRp8CzgIWH6JZpdtH96F_gqL3bXXR3VHzyEZIBKG_wDI9ggh1McIAJjjDBHiZ4gAnuYYIHmGCACfYwwSOYPEaf3x0ev31PYscNosB13BArRVnIissi06WloJKVkipVObWSc2EY09RQWdqUS1oKxaWifCqkLvNMZYbnT9CkOWvMU4TLXJbUZlOtwWXXGeO2NFPLxNQUzKZC76G8f4G1iuXoXVeUVX2d-PYQGa5ah3Isfzmf9bKpo0kZTMUaAHftlc9ueKfn6M72x_ACTTbnnXmJbqsfm5P2_FVE2y9DjZ1g |
| linkProvider | Elsevier |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=IIRNet%3A+A+lightweight+deep+neural+network+using+intensely+inverted+residuals+for+image+recognition&rft.jtitle=Image+and+vision+computing&rft.au=Li%2C+Yuyuan&rft.au=Zhang%2C+Dong&rft.au=Lee%2C+Dah-Jye&rft.date=2019-12-01&rft.issn=0262-8856&rft.volume=92&rft.spage=103819&rft_id=info:doi/10.1016%2Fj.imavis.2019.10.005&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_imavis_2019_10_005 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0262-8856&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0262-8856&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0262-8856&client=summon |