SlimDL: Deploying ultra-light deep learning model on sweeping robots
Advanced object detection methods have yielded impressive progress in recent years. However, the computational constraints of edge mobile devices present significant deployment challenges for state-of-the-art algorithms. We propose a deep learning deployment framework with two stages: model adaptati...
Uloženo v:
| Vydáno v: | Engineering applications of artificial intelligence Ročník 149; s. 110415 |
|---|---|
| Hlavní autoři: | , , , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
Elsevier Ltd
01.06.2025
|
| Témata: | |
| ISSN: | 0952-1976 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Advanced object detection methods have yielded impressive progress in recent years. However, the computational constraints of edge mobile devices present significant deployment challenges for state-of-the-art algorithms. We propose a deep learning deployment framework with two stages: model adaptation and compression. Our method enhance “You Only Look Once version 5” (YOLOv5) with lightweight modules, which improves detection performance while reducing computational load. Additionally, we present a pruning algorithm, employing adaptive batch normalization and iterative pruning. Our evaluation on “Microsoft Common Objects in Context” (MSCOCO) dataset and custom SweepRobot datasets demonstrates that our method consistently outperforms state-of-the-art approaches. On the SweepRobot dataset, our method doubled YOLOv5’s detection speed on the sweeping robot from 15.69 frames per second (FPS) to 30.77 FPS, maintaining 97.3% performance at 20% of the computational cost. Even on Graphics Processing Unit equipped devices, our method achieved 1.8% and 2.8% higher Average Precision compared to direct scaling and pruning with the original pruning algorithm. |
|---|---|
| AbstractList | Advanced object detection methods have yielded impressive progress in recent years. However, the computational constraints of edge mobile devices present significant deployment challenges for state-of-the-art algorithms. We propose a deep learning deployment framework with two stages: model adaptation and compression. Our method enhance “You Only Look Once version 5” (YOLOv5) with lightweight modules, which improves detection performance while reducing computational load. Additionally, we present a pruning algorithm, employing adaptive batch normalization and iterative pruning. Our evaluation on “Microsoft Common Objects in Context” (MSCOCO) dataset and custom SweepRobot datasets demonstrates that our method consistently outperforms state-of-the-art approaches. On the SweepRobot dataset, our method doubled YOLOv5’s detection speed on the sweeping robot from 15.69 frames per second (FPS) to 30.77 FPS, maintaining 97.3% performance at 20% of the computational cost. Even on Graphics Processing Unit equipped devices, our method achieved 1.8% and 2.8% higher Average Precision compared to direct scaling and pruning with the original pruning algorithm. |
| ArticleNumber | 110415 |
| Author | He, Wenbo Sun, Xudong Tong, Chao Wang, Yu Liu, Zhanglin Gao, Shaoxuan |
| Author_xml | – sequence: 1 givenname: Xudong surname: Sun fullname: Sun, Xudong organization: School of Computer Science &, Engineering, Beihang University, Beijing, China – sequence: 2 givenname: Yu surname: Wang fullname: Wang, Yu organization: State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, Beijing, China – sequence: 3 givenname: Zhanglin surname: Liu fullname: Liu, Zhanglin organization: Qfeeltech, Beijing, China – sequence: 4 givenname: Shaoxuan surname: Gao fullname: Gao, Shaoxuan organization: Qfeeltech, Beijing, China – sequence: 5 givenname: Wenbo surname: He fullname: He, Wenbo organization: Department of Computing and Software, McMaster University, Canada – sequence: 6 givenname: Chao orcidid: 0000-0003-4414-4965 surname: Tong fullname: Tong, Chao email: tongchao@buaa.edu.cn organization: School of Computer Science &, Engineering, Beihang University, Beijing, China |
| BookMark | eNqFj8tOwzAQRb0oEm3hF1B-IGHsxGnMCtTykiqxANaWY4-LKzeO7ADq35OosGY10h3dx1mQWRc6JOSKQkGB1tf7Arud6nvlCgaMF5RCRfmMzEFwllOxqs_JIqU9AJRNVc_J5tW7w2Z7k22w9-Houl326Yeocu92H0NmEPvMo4rd9DkEgz4LXZa-R31SYmjDkC7ImVU-4eXvXZL3h_u39VO-fXl8Xt9tc824GPLKWNSAojYVtcwqwUtYIViDrTFWKdbY0jZi7LSMosW2BcW1MNYA57pZlUtSn3J1DClFtLKP7qDiUVKQE7_cyz9-OfHLE_9ovD0ZcVz35TDKpB12Go2LqAdpgvsv4gfgOG16 |
| Cites_doi | 10.1109/CVPR.2017.690 10.1109/TPAMI.2015.2389824 10.1145/3447582 10.1109/ACCESS.2022.3182659 10.1109/CVPR42600.2020.01079 10.1007/s11263-021-01453-z 10.1109/ICCV48922.2021.00052 10.1109/ICCV48922.2021.00447 10.1109/CVPR52729.2023.01544 10.1109/ICCV.2019.00140 10.1109/CVPR52729.2023.01157 10.1007/978-3-030-01264-9_8 10.1109/CVPR52729.2023.00721 10.1109/ICCV.2017.324 10.1109/CVPR.2016.91 10.1109/CVPR.2018.00474 10.1109/CVPR.2018.00716 10.1109/ACCESS.2015.2494536 10.1109/CVPR42600.2020.00165 |
| ContentType | Journal Article |
| Copyright | 2025 Elsevier Ltd |
| Copyright_xml | – notice: 2025 Elsevier Ltd |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.engappai.2025.110415 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Applied Sciences Computer Science |
| ExternalDocumentID | 10_1016_j_engappai_2025_110415 S0952197625004154 |
| GroupedDBID | --K --M .DC .~1 0R~ 1B1 1~. 1~5 4.4 457 4G. 5GY 5VS 7-5 71M 8P~ 9JN AABNK AAEDT AAEDW AAIKJ AAKOC AALRI AAOAW AAQFI AATTM AAXKI AAXUO AAYFN AAYWO ABBOA ABJNI ABMAC ACDAQ ACGFS ACRLP ACVFH ACZNC ADBBV ADCNI ADEZE ADTZH AEBSH AECPX AEIPS AEKER AENEX AEUPX AFJKZ AFPUW AFTJW AGCQF AGHFR AGUBO AGYEJ AHHHB AHJVU AHZHX AIALX AIEXJ AIGII AIIUN AIKHN AITUG AKBMS AKRWK AKYEP ALMA_UNASSIGNED_HOLDINGS AMRAJ ANKPU AOUOD APXCP AXJTR BJAXD BKOJK BLXMC CS3 DU5 EBS EFJIC EFKBS EO8 EO9 EP2 EP3 F5P FDB FIRID FNPLU FYGXN G-Q GBLVA GBOLZ IHE J1W JJJVA KOM LG9 LY7 M41 MO0 N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 ROL RPZ SDF SDG SDP SES SEW SPC SPCBC SST SSV SSZ T5K TN5 ~G- 29G 9DU AAQXK AAYXX ABWVN ABXDB ACLOT ACNNM ACRPL ADJOM ADMUD ADNMO AGQPQ ASPBG AVWKF AZFZN CITATION EFLBG EJD FEDTE FGOYB G-2 HLZ HVGLF HZ~ R2- SBC SET UHS WUQ ZMT ~HD |
| ID | FETCH-LOGICAL-c259t-4dfec0e96d41f2fa95307e0fdebddfaa28f3f89deef21efebb0a5c9dfd055c873 |
| ISICitedReferencesCount | 0 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001445530700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0952-1976 |
| IngestDate | Sat Nov 29 06:57:11 EST 2025 Sat Aug 23 17:11:43 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Model compression Model adaptation Object detection |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c259t-4dfec0e96d41f2fa95307e0fdebddfaa28f3f89deef21efebb0a5c9dfd055c873 |
| ORCID | 0000-0003-4414-4965 |
| ParticipantIDs | crossref_primary_10_1016_j_engappai_2025_110415 elsevier_sciencedirect_doi_10_1016_j_engappai_2025_110415 |
| PublicationCentury | 2000 |
| PublicationDate | 2025-06-01 2025-06-00 |
| PublicationDateYYYYMMDD | 2025-06-01 |
| PublicationDate_xml | – month: 06 year: 2025 text: 2025-06-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationTitle | Engineering applications of artificial intelligence |
| PublicationYear | 2025 |
| Publisher | Elsevier Ltd |
| Publisher_xml | – name: Elsevier Ltd |
| References | Ge, Liu, Wang, Li, Sun (b6) 2021 Howard, Zhu, Chen, Kalenichenko, Wang, Weyand, Andreetto, Adam (b13) 2017 Vadera, Ameen (b36) 2022; 10 Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.-C., 2018. Mobilenetv2: Inverted residuals and linear bottlenecks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 4510–4520. Cai, Shen (b2) 2023 Zhang, X., Zhou, X., Lin, M., Sun, J., 2018. Shufflenet: An extremely efficient convolutional neural network for mobile devices. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 6848–6856. Li, Y., Chen, Y., Dai, X., Chen, D., Liu, M., Yuan, L., Liu, Z., Zhang, L., Vasconcelos, N., 2021. Micronet: Improving image recognition with extremely low flops. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 468–477. Han, Pool, Tran, Dally (b8) 2015; 28 Li, Li, Geng, Jiang, Cheng, Zhang, Ke, Xu, Chu (b17) 2023 Zhou, Hou, Chen, Feng, Yan (b43) 2020 Han, K., Wang, Y., Tian, Q., Guo, J., Xu, C., Xu, C., 2020. Ghostnet: More features from cheap operations. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1580–1589. Howard, A., Sandler, M., Chu, G., Chen, L.-C., Chen, B., Tan, M., Wang, W., Zhu, Y., Pang, R., Vasudevan, V., et al., 2019. Searching for mobilenetv3. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 1314–1324. Renda, Frankle, Carbin (b30) 2020 Tan, Le (b32) 2019 Ren, Xiao, Chang, Huang, Li, Chen, Wang (b29) 2021; 54 Wang, Qin, Bai, Fu (b38) 2023 Redmon, J., Divvala, S., Girshick, R., Farhadi, A., 2016. You only look once: Unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 779–788. Wang, C.-Y., Bochkovskiy, A., Liao, H.-Y.M., 2023a. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 7464–7475. Iandola, Han, Moskewicz, Ashraf, Dally, Keutzer (b14) 2016 Ma, N., Zhang, X., Zheng, H.-T., Sun, J., 2018. Shufflenet v2: Practical guidelines for efficient cnn architecture design. In: Proceedings of the European Conference on Computer Vision. ECCV, pp. 116–131. Redmon, Farhadi (b27) 2018 Liu, Zhang, Kuang, Zhou, Xue, Wang, Chen, Yang, Liao, Zhang (b21) 2021 Luo, Wu (b22) 2017 Tan, Le (b33) 2021 He, Kang, Dong, Fu, Yang (b10) 2018 Krizhevsky, Sutskever, Hinton (b15) 2012; 25 Li, Wu, Su, Wang (b19) 2020 Lin, T.-Y., Goyal, P., Girshick, R., He, K., Dollár, P., 2017. Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision. pp. 2980–2988. Wang, Qin, Zhang, Fu (b39) 2020 Ren, He, Girshick, Sun (b28) 2015; 28 Chen, J., Kao, S.-h., He, H., Zhuo, W., Wen, S., Lee, C.-H., Chan, S.-H.G., 2023. Run, Don’t Walk: Chasing Higher FLOPS for Faster Neural Networks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 12021–12031. Tang, Han, Guo, Xu, Xu, Wang (b35) 2022; 35 Wang, Yeh, Liao (b40) 2021 Redmon, J., Farhadi, A., 2017. YOLO9000: better, faster, stronger. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 7263–7271. Li, Li, Jiang, Weng, Geng, Li, Ke, Li, Cheng, Nie (b18) 2022 Tan, M., Pang, R., Le, Q.V., 2020. Efficientdet: Scalable and efficient object detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10781–10790. Polyak, Wolf (b24) 2015; 3 Zhang, Li, Li, Liu, Xue, Zhang, Jiang, Huang, Wang, Wang (b41) 2023 Fang, G., Ma, X., Song, M., Mi, M.B., Wang, X., 2023. Depgraph: Towards any structural pruning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16091–16101. He, Zhang, Ren, Sun (b11) 2015; 37 Ding, X., Hao, T., Tan, J., Liu, J., Han, J., Guo, Y., Ding, G., 2021. Resrep: Lossless cnn pruning via decoupling remembering and forgetting. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 4510–4520. Bochkovskiy, Wang, Liao (b1) 2020 Gou, Yu, Maybank, Tao (b7) 2021; 129 Han (10.1016/j.engappai.2025.110415_b8) 2015; 28 Bochkovskiy (10.1016/j.engappai.2025.110415_b1) 2020 Wang (10.1016/j.engappai.2025.110415_b39) 2020 Iandola (10.1016/j.engappai.2025.110415_b14) 2016 Krizhevsky (10.1016/j.engappai.2025.110415_b15) 2012; 25 Tan (10.1016/j.engappai.2025.110415_b33) 2021 Li (10.1016/j.engappai.2025.110415_b17) 2023 Li (10.1016/j.engappai.2025.110415_b18) 2022 10.1016/j.engappai.2025.110415_b20 10.1016/j.engappai.2025.110415_b42 10.1016/j.engappai.2025.110415_b23 Tan (10.1016/j.engappai.2025.110415_b32) 2019 Wang (10.1016/j.engappai.2025.110415_b40) 2021 10.1016/j.engappai.2025.110415_b25 10.1016/j.engappai.2025.110415_b26 10.1016/j.engappai.2025.110415_b4 10.1016/j.engappai.2025.110415_b3 Redmon (10.1016/j.engappai.2025.110415_b27) 2018 Ge (10.1016/j.engappai.2025.110415_b6) 2021 10.1016/j.engappai.2025.110415_b5 Gou (10.1016/j.engappai.2025.110415_b7) 2021; 129 Liu (10.1016/j.engappai.2025.110415_b21) 2021 10.1016/j.engappai.2025.110415_b9 Zhou (10.1016/j.engappai.2025.110415_b43) 2020 He (10.1016/j.engappai.2025.110415_b11) 2015; 37 He (10.1016/j.engappai.2025.110415_b10) 2018 Tang (10.1016/j.engappai.2025.110415_b35) 2022; 35 Ren (10.1016/j.engappai.2025.110415_b29) 2021; 54 Vadera (10.1016/j.engappai.2025.110415_b36) 2022; 10 Wang (10.1016/j.engappai.2025.110415_b38) 2023 Cai (10.1016/j.engappai.2025.110415_b2) 2023 10.1016/j.engappai.2025.110415_b31 Zhang (10.1016/j.engappai.2025.110415_b41) 2023 10.1016/j.engappai.2025.110415_b12 10.1016/j.engappai.2025.110415_b34 Howard (10.1016/j.engappai.2025.110415_b13) 2017 Renda (10.1016/j.engappai.2025.110415_b30) 2020 10.1016/j.engappai.2025.110415_b16 Ren (10.1016/j.engappai.2025.110415_b28) 2015; 28 Li (10.1016/j.engappai.2025.110415_b19) 2020 10.1016/j.engappai.2025.110415_b37 Polyak (10.1016/j.engappai.2025.110415_b24) 2015; 3 Luo (10.1016/j.engappai.2025.110415_b22) 2017 |
| References_xml | – year: 2020 ident: b30 article-title: Comparing rewinding and fine-tuning in neural network pruning – reference: Howard, A., Sandler, M., Chu, G., Chen, L.-C., Chen, B., Tan, M., Wang, W., Zhu, Y., Pang, R., Vasudevan, V., et al., 2019. Searching for mobilenetv3. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 1314–1324. – reference: Redmon, J., Divvala, S., Girshick, R., Farhadi, A., 2016. You only look once: Unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 779–788. – year: 2017 ident: b13 article-title: Mobilenets: Efficient convolutional neural networks for mobile vision applications – start-page: 639 year: 2020 end-page: 654 ident: b19 article-title: Eagleeye: Fast sub-net evaluation for efficient neural network pruning publication-title: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part II 16 – year: 2016 ident: b14 article-title: SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and – reference: Li, Y., Chen, Y., Dai, X., Chen, D., Liu, M., Yuan, L., Liu, Z., Zhang, L., Vasconcelos, N., 2021. Micronet: Improving image recognition with extremely low flops. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 468–477. – reference: Tan, M., Pang, R., Le, Q.V., 2020. Efficientdet: Scalable and efficient object detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10781–10790. – volume: 10 start-page: 63280 year: 2022 end-page: 63300 ident: b36 article-title: Methods for pruning deep neural networks publication-title: IEEE Access – volume: 28 year: 2015 ident: b28 article-title: Faster r-cnn: Towards real-time object detection with region proposal networks publication-title: Adv. Neural Inf. Process. Syst. – start-page: 680 year: 2020 end-page: 697 ident: b43 article-title: Rethinking bottleneck structure for efficient mobile network design publication-title: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part III 16 – reference: Redmon, J., Farhadi, A., 2017. YOLO9000: better, faster, stronger. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 7263–7271. – reference: Ding, X., Hao, T., Tan, J., Liu, J., Han, J., Guo, Y., Ding, G., 2021. Resrep: Lossless cnn pruning via decoupling remembering and forgetting. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 4510–4520. – volume: 129 start-page: 1789 year: 2021 end-page: 1819 ident: b7 article-title: Knowledge distillation: A survey publication-title: Int. J. Comput. Vis. – volume: 28 year: 2015 ident: b8 article-title: Learning both weights and connections for efficient neural network publication-title: Adv. Neural Inf. Process. Syst. – reference: Ma, N., Zhang, X., Zheng, H.-T., Sun, J., 2018. Shufflenet v2: Practical guidelines for efficient cnn architecture design. In: Proceedings of the European Conference on Computer Vision. ECCV, pp. 116–131. – start-page: 6105 year: 2019 end-page: 6114 ident: b32 article-title: Efficientnet: Rethinking model scaling for convolutional neural networks publication-title: International Conference on Machine Learning – reference: Wang, C.-Y., Bochkovskiy, A., Liao, H.-Y.M., 2023a. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 7464–7475. – year: 2023 ident: b2 article-title: FalconNet: Factorization for the light-weight ConvNets – reference: Fang, G., Ma, X., Song, M., Mi, M.B., Wang, X., 2023. Depgraph: Towards any structural pruning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16091–16101. – start-page: 10096 year: 2021 end-page: 10106 ident: b33 article-title: Efficientnetv2: Smaller models and faster training publication-title: International Conference on Machine Learning – volume: 25 year: 2012 ident: b15 article-title: Imagenet classification with deep convolutional neural networks publication-title: Adv. Neural Inf. Process. Syst. – year: 2018 ident: b27 article-title: Yolov3: An incremental improvement – reference: Chen, J., Kao, S.-h., He, H., Zhuo, W., Wen, S., Lee, C.-H., Chan, S.-H.G., 2023. Run, Don’t Walk: Chasing Higher FLOPS for Faster Neural Networks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 12021–12031. – year: 2021 ident: b6 article-title: Yolox: Exceeding yolo series in 2021 – reference: Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.-C., 2018. Mobilenetv2: Inverted residuals and linear bottlenecks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 4510–4520. – year: 2021 ident: b40 article-title: You only learn one representation: Unified network for multiple tasks – volume: 35 start-page: 9969 year: 2022 end-page: 9982 ident: b35 article-title: GhostNetv2: enhance cheap operation with long-range attention publication-title: Adv. Neural Inf. Process. Syst. – year: 2020 ident: b1 article-title: Yolov4: Optimal speed and accuracy of object detection – reference: Han, K., Wang, Y., Tian, Q., Guo, J., Xu, C., Xu, C., 2020. Ghostnet: More features from cheap operations. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1580–1589. – year: 2020 ident: b39 article-title: Neural pruning via growing regularization – year: 2022 ident: b18 article-title: YOLOv6: A single-stage object detection framework for industrial applications – reference: Lin, T.-Y., Goyal, P., Girshick, R., He, K., Dollár, P., 2017. Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision. pp. 2980–2988. – start-page: 7021 year: 2021 end-page: 7032 ident: b21 article-title: Group fisher pruning for practical network compression publication-title: International Conference on Machine Learning – year: 2023 ident: b41 article-title: Rethinking mobile block for efficient neural models – year: 2023 ident: b17 article-title: Yolov6 v3. 0: A full-scale reloading – volume: 54 start-page: 1 year: 2021 end-page: 34 ident: b29 article-title: A comprehensive survey of neural architecture search: Challenges and solutions publication-title: ACM Comput. Surv. – reference: Zhang, X., Zhou, X., Lin, M., Sun, J., 2018. Shufflenet: An extremely efficient convolutional neural network for mobile devices. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp. 6848–6856. – volume: 3 start-page: 2163 year: 2015 end-page: 2175 ident: b24 article-title: Channel-level acceleration of deep face representations publication-title: IEEE Access – year: 2018 ident: b10 article-title: Soft filter pruning for accelerating deep convolutional neural networks – volume: 37 start-page: 1904 year: 2015 end-page: 1916 ident: b11 article-title: Spatial pyramid pooling in deep convolutional networks for visual recognition publication-title: IEEE Trans. Pattern Anal. Mach. Intell. – year: 2017 ident: b22 article-title: An entropy-based pruning method for cnn compression – year: 2023 ident: b38 article-title: Why is the state of neural network pruning so confusing? on the fairness, comparison setup, and trainability in network pruning – year: 2021 ident: 10.1016/j.engappai.2025.110415_b40 – start-page: 680 year: 2020 ident: 10.1016/j.engappai.2025.110415_b43 article-title: Rethinking bottleneck structure for efficient mobile network design – year: 2020 ident: 10.1016/j.engappai.2025.110415_b39 – ident: 10.1016/j.engappai.2025.110415_b26 doi: 10.1109/CVPR.2017.690 – year: 2017 ident: 10.1016/j.engappai.2025.110415_b22 – volume: 35 start-page: 9969 year: 2022 ident: 10.1016/j.engappai.2025.110415_b35 article-title: GhostNetv2: enhance cheap operation with long-range attention publication-title: Adv. Neural Inf. Process. Syst. – year: 2023 ident: 10.1016/j.engappai.2025.110415_b38 – start-page: 639 year: 2020 ident: 10.1016/j.engappai.2025.110415_b19 article-title: Eagleeye: Fast sub-net evaluation for efficient neural network pruning – volume: 25 year: 2012 ident: 10.1016/j.engappai.2025.110415_b15 article-title: Imagenet classification with deep convolutional neural networks publication-title: Adv. Neural Inf. Process. Syst. – year: 2023 ident: 10.1016/j.engappai.2025.110415_b17 – volume: 37 start-page: 1904 issue: 9 year: 2015 ident: 10.1016/j.engappai.2025.110415_b11 article-title: Spatial pyramid pooling in deep convolutional networks for visual recognition publication-title: IEEE Trans. Pattern Anal. Mach. Intell. doi: 10.1109/TPAMI.2015.2389824 – volume: 54 start-page: 1 issue: 4 year: 2021 ident: 10.1016/j.engappai.2025.110415_b29 article-title: A comprehensive survey of neural architecture search: Challenges and solutions publication-title: ACM Comput. Surv. doi: 10.1145/3447582 – start-page: 7021 year: 2021 ident: 10.1016/j.engappai.2025.110415_b21 article-title: Group fisher pruning for practical network compression – volume: 10 start-page: 63280 year: 2022 ident: 10.1016/j.engappai.2025.110415_b36 article-title: Methods for pruning deep neural networks publication-title: IEEE Access doi: 10.1109/ACCESS.2022.3182659 – year: 2022 ident: 10.1016/j.engappai.2025.110415_b18 – ident: 10.1016/j.engappai.2025.110415_b34 doi: 10.1109/CVPR42600.2020.01079 – volume: 129 start-page: 1789 year: 2021 ident: 10.1016/j.engappai.2025.110415_b7 article-title: Knowledge distillation: A survey publication-title: Int. J. Comput. Vis. doi: 10.1007/s11263-021-01453-z – ident: 10.1016/j.engappai.2025.110415_b16 doi: 10.1109/ICCV48922.2021.00052 – volume: 28 year: 2015 ident: 10.1016/j.engappai.2025.110415_b8 article-title: Learning both weights and connections for efficient neural network publication-title: Adv. Neural Inf. Process. Syst. – ident: 10.1016/j.engappai.2025.110415_b4 doi: 10.1109/ICCV48922.2021.00447 – ident: 10.1016/j.engappai.2025.110415_b5 doi: 10.1109/CVPR52729.2023.01544 – ident: 10.1016/j.engappai.2025.110415_b12 doi: 10.1109/ICCV.2019.00140 – ident: 10.1016/j.engappai.2025.110415_b3 doi: 10.1109/CVPR52729.2023.01157 – ident: 10.1016/j.engappai.2025.110415_b23 doi: 10.1007/978-3-030-01264-9_8 – start-page: 10096 year: 2021 ident: 10.1016/j.engappai.2025.110415_b33 article-title: Efficientnetv2: Smaller models and faster training – ident: 10.1016/j.engappai.2025.110415_b37 doi: 10.1109/CVPR52729.2023.00721 – ident: 10.1016/j.engappai.2025.110415_b20 doi: 10.1109/ICCV.2017.324 – year: 2020 ident: 10.1016/j.engappai.2025.110415_b1 – year: 2017 ident: 10.1016/j.engappai.2025.110415_b13 – year: 2021 ident: 10.1016/j.engappai.2025.110415_b6 – year: 2018 ident: 10.1016/j.engappai.2025.110415_b10 – year: 2016 ident: 10.1016/j.engappai.2025.110415_b14 – year: 2020 ident: 10.1016/j.engappai.2025.110415_b30 – start-page: 6105 year: 2019 ident: 10.1016/j.engappai.2025.110415_b32 article-title: Efficientnet: Rethinking model scaling for convolutional neural networks – ident: 10.1016/j.engappai.2025.110415_b25 doi: 10.1109/CVPR.2016.91 – ident: 10.1016/j.engappai.2025.110415_b31 doi: 10.1109/CVPR.2018.00474 – year: 2018 ident: 10.1016/j.engappai.2025.110415_b27 – ident: 10.1016/j.engappai.2025.110415_b42 doi: 10.1109/CVPR.2018.00716 – year: 2023 ident: 10.1016/j.engappai.2025.110415_b2 – volume: 3 start-page: 2163 year: 2015 ident: 10.1016/j.engappai.2025.110415_b24 article-title: Channel-level acceleration of deep face representations publication-title: IEEE Access doi: 10.1109/ACCESS.2015.2494536 – ident: 10.1016/j.engappai.2025.110415_b9 doi: 10.1109/CVPR42600.2020.00165 – year: 2023 ident: 10.1016/j.engappai.2025.110415_b41 – volume: 28 year: 2015 ident: 10.1016/j.engappai.2025.110415_b28 article-title: Faster r-cnn: Towards real-time object detection with region proposal networks publication-title: Adv. Neural Inf. Process. Syst. |
| SSID | ssj0003846 |
| Score | 2.433313 |
| Snippet | Advanced object detection methods have yielded impressive progress in recent years. However, the computational constraints of edge mobile devices present... |
| SourceID | crossref elsevier |
| SourceType | Index Database Publisher |
| StartPage | 110415 |
| SubjectTerms | Model adaptation Model compression Object detection |
| Title | SlimDL: Deploying ultra-light deep learning model on sweeping robots |
| URI | https://dx.doi.org/10.1016/j.engappai.2025.110415 |
| Volume | 149 |
| WOSCitedRecordID | wos001445530700001&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 issn: 0952-1976 databaseCode: AIEXJ dateStart: 19950201 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.sciencedirect.com omitProxy: false ssIdentifier: ssj0003846 providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9QwELaWLQcupbxEC0U-cKtc8mzs3iq2vFRVSFvQciFy_Cip0mSV3ZT9-fiZRFAJEOISRaPE3p35NJkZzwOAl1Jiyngmke7mhpT_FaMio0coCXnIOJYyNaHsz2fZ-TleLMjHyeSrr4W5qbK6xpsNWf5XUSuaErYunf0LcfeLKoK6V0JXVyV2df0jwc-r8np2pj39mdDTfHUsoKvWLUWVdsQPuBBLPyzi0k7C0ScGq--KriltUzS2vVMfsR96Fh6MD7xNDkFrko3M6I9Rd8_hrMlotUXHG_eJNMF7q2C-dH06UNmZUxIdva7KHq9vqQnkzr_RZtM5GLsIRZQOmVQ2bOZLZ4Y8JRt_jFBI7PCXXhXb9qW_qHUbYbg6FPWl-pu0PNTb6AqGxNaC_tQye64X12sr-04_ktwBW1GWEjwFWyfvTxcf-m91jG0pl_8xoxry23e73XwZmSQXO2Db-RLwxGLgAZiI-iG47_wK6LT2SpH86A5PewRmFiXHsMcIHGEEaoxAjxFoMAKbGnqMQIuRx-DTm9OL1--QG6iBmPJy1yjhUrBAkCOehDKSlKRKw4tAclFwLimNsIwlJmoPGYVCiqIIaMoIlzxIU4az-AmY1k0tngIowjjWc5sEi4MkyXhBE1ZQogzCCHOc8l3wyvMpX9q-KblPKLzKPWdzzdnccnYXEM_O3Fl_1qrLFQp-8-7eP7z7DNwbQPscTNdtJ_bBXXazLlftCweYH_8phlM |
| 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=SlimDL%3A+Deploying+ultra-light+deep+learning+model+on+sweeping+robots&rft.jtitle=Engineering+applications+of+artificial+intelligence&rft.au=Sun%2C+Xudong&rft.au=Wang%2C+Yu&rft.au=Liu%2C+Zhanglin&rft.au=Gao%2C+Shaoxuan&rft.date=2025-06-01&rft.pub=Elsevier+Ltd&rft.issn=0952-1976&rft.volume=149&rft_id=info:doi/10.1016%2Fj.engappai.2025.110415&rft.externalDocID=S0952197625004154 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0952-1976&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0952-1976&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0952-1976&client=summon |