Abstract
Adaptive optimization methods for deep learning adjust the inference task to the current circumstances at runtime to improve the resource footprint while maintaining the model’s performance. These methods are essential for the widespread adoption of deep learning, as they offer a way to reduce the resource footprint of the inference task while also having access to additional information about the current environment. This survey covers the state-of-the-art at-runtime optimization methods, provides guidance for readers to choose the best method for their specific use-case, and also highlights current research gaps in this field.
- [1] . 2018. SnaPEA: Predictive early activation for reducing computation in deep convolutional neural networks. In 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA). 662–673.
DOI: Google ScholarDigital Library - [2] . 2016. Dynamic capacity networks. In Proceedings of The 33rd International Conference on Machine Learning. PMLR, 2549–2558.Google Scholar
- [3] . 2018. Action search: Spotting actions in videos and its application to temporal action localization. In Proceedings of the European Conference on Computer Vision (ECCV). 251–266.Google ScholarDigital Library
- [4] . 2021. TempDiff: Temporal difference-based feature map-level sparsity induction in CNNs with \(\lt\)4% memory overhead. In 2021 IEEE 3rd International Conference on Artificial Intelligence Circuits and Systems (AICAS). 1–4.
DOI: Google ScholarCross Ref - [5] . 2016. Impatient DNNs - Deep Neural Networks with Dynamic Time Budgets.
DOI: arxiv:1610.02850 [cs].Google ScholarCross Ref - [6] . 2019. Frugal following: Power thrifty object detection and tracking for mobile augmented reality. In Proceedings of the 17th Conference on Embedded Networked Sensor Systems (SenSys’19). Association for Computing Machinery, New York, NY, USA, 96–109.
DOI: Google ScholarDigital Library - [7] . 2020. Batch-Shaping for Learning Conditional Channel Gated Networks.
DOI: arxiv:1907.06627 [cs, stat].Google ScholarCross Ref - [8] . 2016. Conditional Computation in Neural Networks for Faster Models.
DOI: arxiv:1511.06297 [cs].Google ScholarCross Ref - [9] . 2017. Adaptive neural networks for efficient inference. In Proceedings of the 34th International Conference on Machine Learning. PMLR, 527–536.Google ScholarDigital Library
- [10] . 2017. Adaptive neural networks for fast test-time prediction. (
Feb. 2017).Google Scholar - [11] . 2018. EVA\(^2\): Exploiting temporal redundancy in live computer vision. In 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture (ISCA). 533–546.
DOI: Google ScholarDigital Library - [12] . 2022. Dynamic pruning for parsimonious CNN inference on embedded systems. In Design and Architecture for Signal and Image Processing (Lecture Notes in Computer Science), and (Eds.). Springer International Publishing, Cham, 45–56.
DOI: Google ScholarDigital Library - [13] . 2021. Dynamic routing networks. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 3588–3597.Google ScholarCross Ref
- [14] . 2018. Skip RNN: Learning to Skip State Updates in Recurrent Neural Networks.
DOI: arxiv:1708.06834 [cs].Google ScholarCross Ref - [15] . 2019. Scaling video analytics on constrained edge nodes. Proceedings of Machine Learning and Systems 1 (
April 2019), 406–417.Google Scholar - [16] . 2019. SeerNet: Predicting convolutional neural network feature-map sparsity through low-bit quantization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 11216–11225.Google ScholarCross Ref
- [17] . 2020. CBinfer: Exploiting frame-to-frame locality for faster convolutional network inference on video streams. IEEE Transactions on Circuits and Systems for Video Technology 30, 5 (
May 2020), 1451–1465.DOI: Google ScholarCross Ref - [18] . 2017. CBinfer: Change-based inference for convolutional neural networks on video data. In Proceedings of the 11th International Conference on Distributed Smart Cameras (ICDSC 2017). Association for Computing Machinery, New York, NY, USA, 1–8.
DOI: Google ScholarDigital Library - [19] . 2019. Self-adaptive network pruning. In Neural Information Processing (Lecture Notes in Computer Science), , , and (Eds.). Springer International Publishing, Cham, 175–186.
DOI: Google ScholarDigital Library - [20] . 2023. Survey: Exploiting data redundancy for optimization of deep learning. Comput. Surveys 55, 10 (
Feb. 2023), 212:1–212:38.DOI: Google ScholarDigital Library - [21] . 2015. Glimpse: Continuous, real-time object recognition on mobile devices. In Proceedings of the 13th ACM Conference on Embedded Networked Sensor Systems (SenSys’15). Association for Computing Machinery, New York, NY, USA, 155–168.
DOI: Google ScholarDigital Library - [22] . 2020. Dynamic ReLU. In Computer Vision – ECCV 2020 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 351–367.
DOI: Google ScholarDigital Library - [23] . 2019. You look twice: GaterNet for dynamic filter selection in CNNs. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 9172–9180.Google ScholarCross Ref
- [24] . 2020. InstaNAS: Instance-aware neural architecture search. Proceedings of the AAAI Conference on Artificial Intelligence 34, 04 (
April 2020), 3577–3584.DOI: Google ScholarCross Ref - [25] . 2021. Optimal branch location for cost-effective inference on Branchynet. In 2021 IEEE International Conference on Big Data (Big Data). 5071–5080.
DOI: Google ScholarCross Ref - [26] . 2009. Similarity caching. In Proceedings of the Twenty-Eighth ACM SIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems (PODS’09). Association for Computing Machinery, New York, NY, USA, 127–136.
DOI: Google ScholarDigital Library - [27] . 2021. Differentiable patch selection for image recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2351–2360.Google ScholarCross Ref
- [28] . 2022. Memory-aware and context-aware multi-DNN inference on the edge. Pervasive and Mobile Computing 83 (
July 2022), 101594.DOI: Google ScholarDigital Library - [29] . 2022. A two-step machine learning approach for dynamic model selection: A case study on a micro milling process. Computers in Industry 143 (
Dec. 2022), 103764.DOI: Google ScholarDigital Library - [30] . 2019. Universal Transformers.
DOI: arxiv:1807.03819 [cs, stat].Google ScholarCross Ref - [31] . 2017. More is less: A more complicated network with less inference complexity. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 5840–5848.Google ScholarCross Ref
- [32] Utsav Drolia, Katherine Guo, and Priya Narasimhan. 2017. Precog: Prefetching for image recognition applications at the edge. In Proceedings of the Second ACM/IEEE Symposium on Edge Computing (SEC’17). Association for Computing Machinery, New York, NY, USA, 1–13. Google ScholarDigital Library
- [33] . 2017. Cachier: Edge-caching for recognition applications. In 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS). 276–286.
DOI: Google ScholarCross Ref - [34] . 2020. Dynamic channel and layer gating in convolutional neural networks. In KI 2020: Advances in Artificial Intelligence (Lecture Notes in Computer Science), , , and (Eds.). Springer International Publishing, Cham, 33–45.
DOI: Google ScholarDigital Library - [35] . 2020. Depth-Adaptive Transformer.
DOI: arxiv:1910.10073 [cs].Google ScholarCross Ref - [36] . 2022. Fire together wire together: A dynamic pruning approach with self-supervised mask prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 12454–12463.Google ScholarCross Ref
- [37] 2023. GPT-4 Technical Report.
Technical Report arXiv:2303.08774.DOI: arxiv:2303.08774 [cs].Google ScholarCross Ref - [38] . 2008. A metric cache for similarity search. In Proceedings of the 2008 ACM Workshop on Large-Scale Distributed Systems for Information Retrieval (LSDS-IR’08). Association for Computing Machinery, New York, NY, USA, 43–50.
DOI: Google ScholarDigital Library - [39] . 2009. Caching content-based queries for robust and efficient image retrieval. In Proceedings of the 12th International Conference on Extending Database Technology: Advances in Database Technology (EDBT’09). Association for Computing Machinery, New York, NY, USA, 780–790.
DOI: Google ScholarDigital Library - [40] . 2012. Similarity caching in large-scale image retrieval. Information Processing & Management 48, 5 (
Sept. 2012), 803–818.DOI: Google ScholarDigital Library - [41] . 2018. Watching a Small Portion Could Be as Good as Watching All: Towards Efficient Video Classification.Google Scholar
- [42] . 2020. FlexDNN: Input-adaptive on-device deep learning for efficient mobile vision. In 2020 IEEE/ACM Symposium on Edge Computing (SEC). 84–95.
DOI: Google ScholarCross Ref - [43] . 2020. CacheNet: A Model Caching Framework for Deep Learning Inference on the Edge.
DOI: arxiv:2007.01793 [cs, eess].Google ScholarCross Ref - [44] . 2022. Adaptive token sampling for efficient vision transformers. In Computer Vision – ECCV 2022 (Lecture Notes in Computer Science), , , , , and (Eds.). Springer Nature Switzerland, Cham, 396–414.
DOI: Google ScholarDigital Library - [45] . 2017. Spatially adaptive computation time for residual networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 1039–1048.Google ScholarCross Ref
- [46] . 2022. Accelerating deep learning classification with error-controlled approximate-key caching. In IEEE INFOCOM 2022 - IEEE Conference on Computer Communications. 2118–2127.
DOI: Google ScholarDigital Library - [47] . 2017. Look closer to see better: Recurrent attention convolutional neural network for fine-grained image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 4438–4446.Google ScholarCross Ref
- [48] . 2018. Speed reading: Learning to read forbackward via shuttle. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, Brussels, Belgium, 4439–4448.
DOI: Google ScholarCross Ref - [49] . 2019. Dynamic Channel Pruning: Feature Boosting and Suppression.
DOI: arxiv:1810.05331 [cs].Google ScholarCross Ref - [50] . 2022. T-RECX: Tiny-Resource Efficient Convolutional Neural Networks with Early-Exit.
DOI: arxiv:2207.06613 [cs, eess].Google ScholarCross Ref - [51] . 2021. FrameExit: Conditional early exiting for efficient video recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 15608–15618.Google ScholarCross Ref
- [52] . 2019. Challenges and opportunities of DNN model execution caching. In Proceedings of the Workshop on Distributed Infrastructures for Deep Learning (DIDL’19). Association for Computing Machinery, New York, NY, USA, 7–12.
DOI: Google ScholarDigital Library - [53] . 2020. Intelligent cooperative edge computing in internet of things. IEEE Internet of Things Journal 7, 10 (
Oct. 2020), 9372–9382.DOI: Google ScholarCross Ref - [54] . 2022. Adaptive Sparse Transformer for Multilingual Translation.
DOI: arxiv:2104.07358 [cs].Google ScholarCross Ref - [55] . 2017. Adaptive Computation Time for Recurrent Neural Networks.
DOI: arxiv:1603.08983 [cs].Google ScholarCross Ref - [56] . 2018. Potluck: Cross-application approximate deduplication for computation-intensive mobile applications. In Proceedings of the Twenty-Third International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS’18). Association for Computing Machinery, New York, NY, USA, 271–284.
DOI: Google ScholarDigital Library - [57] . 2018. FoggyCache: Cross-device approximate computation reuse. Living on the Edge (2018), 16.Google Scholar
- [58] . 2019. Dynamic recursive neural network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 5147–5156.Google ScholarCross Ref
- [59] . 2018. A Survey on Methods and Theories of Quantized Neural Networks.
DOI: arxiv:1808.04752 [cs, stat].Google ScholarCross Ref - [60] . 2021. Exploration of block-wise dynamic sparseness. Pattern Recognition Letters 151 (
Nov. 2021), 187–192.DOI: Google ScholarDigital Library - [61] . 2016. MCDNN: An approximation-based execution framework for deep stream processing under resource constraints. In Proceedings of the 14th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys’16). Association for Computing Machinery, New York, NY, USA, 123–136.
DOI: Google ScholarDigital Library - [62] . 2021. Dynamic neural networks: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence (2021), 1–1.
DOI: Google ScholarCross Ref - [63] . 2019. Neural Speed Reading with Structural-Jump-LSTM.
DOI: arxiv:1904.00761 [cs, stat].Google ScholarCross Ref - [64] . 2020. ILFO: Adversarial attack on adaptive neural networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 14264–14273.Google ScholarCross Ref
- [65] . 2020. The tree ensemble layer: Differentiability meets conditional computation. In Proceedings of the 37th International Conference on Machine Learning. PMLR, 4138–4148.Google Scholar
- [66] . 2020. Channel selection using Gumbel Softmax. In Computer Vision – ECCV 2020 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 241–257.
DOI: Google ScholarDigital Library - [67] . 2021. A Panda? No, It’s a Sloth: Slowdown Attacks on Adaptive Multi-Exit Neural Network Inference.
DOI: arxiv:2010.02432 [cs].Google ScholarCross Ref - [68] . 2020. DynaBERT: Dynamic BERT with adaptive width and depth. In Advances in Neural Information Processing Systems, Vol. 33. Curran Associates, Inc., 9782–9793.Google Scholar
- [69] . 2018. Focus: Querying large video datasets with low latency and low cost. In 13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18). 269–286.Google Scholar
- [70] . 2018. Learning Anytime Predictions in Neural Networks via Adaptive Loss Balancing.
DOI: arxiv:1708.06832 [cs].Google ScholarCross Ref - [71] . 2020. Triple Wins: Boosting Accuracy, Robustness and Efficiency Together by Enabling Input-Adaptive Inference.
DOI: arxiv:2002.10025 [cs].Google ScholarCross Ref - [72] . 2018. Deep learning for image-based cancer detection and diagnosis - a survey. Pattern Recognition 83 (
Nov. 2018), 134–149.DOI: Google ScholarDigital Library - [73] . 2019. Boosting the performance of CNN accelerators with dynamic fine-grained channel gating. In Proceedings of the 52nd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO’52). Association for Computing Machinery, New York, NY, USA, 139–150.
DOI: Google ScholarDigital Library - [74] . 2019. Channel gating neural networks. In Advances in Neural Information Processing Systems, Vol. 32. Curran Associates, Inc.Google Scholar
- [75] . 2018. Multi-Scale Dense Networks for Resource Efficient Image Classification.
DOI: arxiv:1703.09844 [cs].Google ScholarCross Ref - [76] . 2022. Glance and Focus Networks for Dynamic Visual Recognition.
DOI: arxiv:2201.03014 [cs].Google ScholarCross Ref - [77] . 2017. Length adaptive recurrent model for text classification. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management (CIKM’17). Association for Computing Machinery, New York, NY, USA, 1019–1027.
DOI: Google ScholarDigital Library - [78] . 2017. DeepMon: Mobile GPU-based deep learning framework for continuous vision applications. In Proceedings of the 15th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys’17). Association for Computing Machinery, New York, NY, USA, 82–95.
DOI: Google ScholarDigital Library - [79] . 2016. Decision Forests, Convolutional Networks and the Models in-Between.
DOI: arxiv:1603.01250 [cs].Google ScholarCross Ref - [80] . 2020. Spatula: Efficient cross-camera video analytics on large camera networks. In 2020 IEEE/ACM Symposium on Edge Computing (SEC). 110–124.
DOI: Google ScholarCross Ref - [81] . 2019. ReXCam: Resource-Efficient, Cross-Camera Video Analytics at Scale.
DOI: arxiv:1811.01268 [cs].Google ScholarCross Ref - [82] . 2017. Variable Computation in Recurrent Neural Networks.
DOI: arxiv:1611.06188 [cs, stat].Google ScholarCross Ref - [83] . 2018. Chameleon: Scalable adaptation of video analytics. In Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication (SIGCOMM’18). Association for Computing Machinery, New York, NY, USA, 253–266.
DOI: Google ScholarDigital Library - [84] . 2021. Dynamic Slimmable Denoising Network.
DOI: arxiv:2110.08940 [cs, eess].Google ScholarCross Ref - [85] . 2020. AdaBits: Neural network quantization with adaptive bit-widths. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2146–2156.Google ScholarCross Ref
- [86] . 2017. NoScope: Optimizing Neural Network Queries over Video at Scale.
DOI: arxiv:1703.02529 [cs].Google ScholarCross Ref - [87] . 2019. Shallow-deep networks: Understanding and mitigating network overthinking. In Proceedings of the 36th International Conference on Machine Learning. PMLR, 3301–3310.Google Scholar
- [88] . 2021. Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search.
DOI: arxiv:2010.07003 [cs].Google ScholarCross Ref - [89] . 2020. PointRend: Image segmentation as rendering. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 9799–9808.Google ScholarCross Ref
- [90] . 2019. Pixel-wise attentional gating for scene parsing. In 2019 IEEE Winter Conference on Applications of Computer Vision (WACV). 1024–1033.
DOI: Google ScholarCross Ref - [91] . 2015. Deep neural decision forests. In Proceedings of the IEEE International Conference on Computer Vision. 1467–1475.Google ScholarDigital Library
- [92] . 2022. Multi-Exit Semantic Segmentation Networks.
arxiv:2106.03527 [cs].Google Scholar - [93] . 2022. Dynamic Channel Selection in Self-Supervised Learning.
DOI: arxiv:2207.12065 [cs].Google ScholarCross Ref - [94] . 2018. Stochastic downsampling for cost-adjustable inference and improved regularization in convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 7929–7938.Google ScholarCross Ref
- [95] . 2021. Adaptive inference through early-exit networks: Design, challenges and directions. In Proceedings of the 5th International Workshop on Embedded and Mobile Deep Learning (EMDL’21). Association for Computing Machinery, New York, NY, USA, 1–6.
DOI: Google ScholarDigital Library - [96] . 2020. SPINN: Synergistic progressive inference of neural networks over device and cloud. In Proceedings of the 26th Annual International Conference on Mobile Computing and Networking (MobiCom’20). Association for Computing Machinery, New York, NY, USA, 1–15.
DOI: Google ScholarDigital Library - [97] . 2020. Performance analysis of local exit for distributed deep neural networks over cloud and edge computing. ETRI Journal 42, 5 (2020), 658–668.
DOI: Google ScholarCross Ref - [98] . 2018. Anytime Neural Prediction via Slicing Networks Vertically.
DOI: arxiv:1807.02609 [cs, stat].Google ScholarCross Ref - [99] . 2019. MobiSR: Efficient on-device super-resolution through heterogeneous mobile processors. In The 25th Annual International Conference on Mobile Computing and Networking (MobiCom’19). Association for Computing Machinery, New York, NY, USA, 1–16.
DOI: Google ScholarDigital Library - [100] . 2016. Lazy Evaluation of Convolutional Filters.
DOI: arxiv:1605.08543 [cs].Google ScholarCross Ref - [101] . 2017. The cascading neural network: Building the internet of smart things. Knowledge and Information Systems 52, 3 (
Sept. 2017), 791–814.DOI: Google ScholarDigital Library - [102] . 2018. IamNN: Iterative and Adaptive Mobile Neural Network for Efficient Image Classification.
DOI: arxiv:1804.10123 [cs].Google ScholarCross Ref - [103] . 2021. DS-Net++: Dynamic Weight Slicing for Efficient Inference in CNNs and Transformers.
DOI: arxiv:2109.10060 [cs].Google ScholarCross Ref - [104] . 2021. Dynamic slimmable network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8607–8617.Google ScholarCross Ref
- [105] . 2021. 2D or not 2D? Adaptive 3D convolution selection for efficient video recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 6155–6164.Google ScholarCross Ref
- [106] . 2019. Improved techniques for training adaptive deep networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 1891–1900.Google ScholarCross Ref
- [107] . 2018. Deep learning for smart industry: Efficient manufacture inspection system with fog computing. IEEE Transactions on Industrial Informatics 14, 10 (
Oct. 2018), 4665–4673.DOI: Google ScholarCross Ref - [108] . 2017. Not all pixels are equal: Difficulty-aware semantic segmentation via deep layer cascade. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 3193–3202.Google ScholarCross Ref
- [109] . 2020. Learning dynamic routing for semantic segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8553–8562.Google ScholarCross Ref
- [110] . 2017. Dynamic computational time for visual attention. In Proceedings of the IEEE International Conference on Computer Vision Workshops. 1199–1209.Google ScholarCross Ref
- [111] . 2015. Starfish: Efficient concurrency support for computer vision applications. In Proceedings of the 13th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys’15). Association for Computing Machinery, New York, NY, USA, 213–226.
DOI: Google ScholarDigital Library - [112] . 2017. Runtime neural pruning. In Advances in Neural Information Processing Systems, Vol. 30. Curran Associates, Inc.Google Scholar
- [113] . 2017. PredictiveNet: An energy-efficient convolutional neural network via zero prediction. In 2017 IEEE International Symposium on Circuits and Systems (ISCAS). 1–4.
DOI: Google ScholarCross Ref - [114] . 2019. Learning Instance-wise Sparsity for Accelerating Deep Models.
DOI: arxiv:1907.11840 [cs].Google ScholarCross Ref - [115] . 2018. Dynamic deep neural networks: Optimizing accuracy-efficiency trade-offs by selective execution. Proceedings of the AAAI Conference on Artificial Intelligence 32, 1 (
April 2018).DOI: Google ScholarCross Ref - [116] . 2019. Edge assisted real-time object detection for mobile augmented reality. In The 25th Annual International Conference on Mobile Computing and Networking (MobiCom’19). Association for Computing Machinery, New York, NY, USA, 1–16.
DOI: Google ScholarDigital Library - [117] . 2020. Continuous, real-time object detection on mobile devices without offloading. In 2020 IEEE 40th International Conference on Distributed Computing Systems (ICDCS). 976–986.
DOI: Google ScholarCross Ref - [118] . 2018. On-demand deep model compression for mobile devices: A usage-driven model selection framework. In Proceedings of the 16th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys’18). Association for Computing Machinery, New York, NY, USA, 389–400.
DOI: Google ScholarDigital Library - [119] . 2020. FastBERT: A Self-distilling BERT with Adaptive Inference Time.
DOI: arxiv:2004.02178 [cs].Google ScholarCross Ref - [120] . 2020. Finding decision jumps in text classification. Neurocomputing 371 (
Jan. 2020), 177–187.DOI: Google ScholarDigital Library - [121] . 2017. A dynamic deep neural network design for efficient workload allocation in edge computing. In 2017 IEEE International Conference on Computer Design (ICCD). 273–280.
DOI: Google ScholarCross Ref - [122] . 2021. Dynamic-OFA: Runtime DNN architecture switching for performance scaling on heterogeneous embedded platforms. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 3110–3118.Google ScholarCross Ref
- [123] . 2018. CIRCE: Real-time caching for instance recognition on cloud environments and multi-core architectures. In Proceedings of the 26th ACM International Conference on Multimedia (MM’18). Association for Computing Machinery, New York, NY, USA, 346–354.
DOI: Google ScholarDigital Library - [124] . 2022. Toward efficient and adaptive design of video detection system with deep neural networks. ACM Transactions on Embedded Computing Systems 21, 3 (
July 2022), 33:1–33:21.DOI: Google ScholarDigital Library - [125] . 2020. Optimizing deep learning inference on embedded systems through adaptive model selection. ACM Transactions on Embedded Computing Systems 19, 1 (
Feb. 2020), 2:1–2:28.DOI: Google ScholarDigital Library - [126] . 2022. Split computing and early exiting for deep learning applications: Survey and research challenges. Comput. Surveys 55, 5 (
Dec. 2022), 90:1–90:30.DOI: Google ScholarDigital Library - [127] . 2017. Deciding how to decide: Dynamic routing in artificial neural networks. In Proceedings of the 34th International Conference on Machine Learning. PMLR, 2363–2372.Google Scholar
- [128] . 2022. AdaViT: Adaptive vision transformers for efficient image recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 12309–12318.Google ScholarCross Ref
- [129] . 2020. AR-Net: Adaptive frame resolution for efficient action recognition. In Computer Vision – ECCV 2020 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 86–104.
DOI: Google ScholarDigital Library - [130] . 2021. AdaFuse: Adaptive Temporal Fusion Network for Efficient Action Recognition.
DOI: arxiv:2102.05775 [cs].Google ScholarCross Ref - [131] . 2014. Recurrent models of visual attention. In Advances in Neural Information Processing Systems, Vol. 27. Curran Associates, Inc.Google ScholarDigital Library
- [132] . 2018. HydraNets: Specialized dynamic architectures for efficient inference. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 8080–8089.Google Scholar
- [133] . 2022. DeepScale: Online frame size adaptation for multi-object tracking on smart cameras and edge servers. In 2022 IEEE/ACM Seventh International Conference on Internet-of-Things Design and Implementation (IoTDI). 67–79.
DOI: Google ScholarCross Ref - [134] . 2020. Review—Deep learning methods for sensor based predictive maintenance and future perspectives for electrochemical sensors. Journal of The Electrochemical Society 167, 3 (
Jan. 2020), 037552.DOI: Google ScholarCross Ref - [135] . 2016. Learning to Reason with Adaptive Computation.
DOI: arxiv:1610.07647 [cs, stat].Google ScholarCross Ref - [136] . 2016. Sigma Delta Quantized Networks.
DOI: arxiv:1611.02024 [cs].Google ScholarCross Ref - [137] . 2017. Changing Model Behavior at Test-Time Using Reinforcement Learning.
DOI: arxiv:1702.07780 [cs, stat].Google ScholarCross Ref - [138] . 2018. {MODI}: Mobile deep inference made efficient by edge computing. In USENIX Workshop on Hot Topics in Edge Computing (HotEdge 18).Google Scholar
- [139] . 2018. Recurrent residual module for fast inference in videos. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 1536–1545.Google ScholarCross Ref
- [140] . 2021. VA-RED$\(\hat2\)$: Video Adaptive Redundancy Reduction.
DOI: arxiv:2102.07887 [cs].Google ScholarCross Ref - [141] . 2021. IA-RED2: Interpretability-aware redundancy reduction for vision transformers. In Advances in Neural Information Processing Systems, Vol. 34. Curran Associates, Inc., 24898–24911.Google Scholar
- [142] . 2017. FALCON: Feature driven selective classification for energy-efficient image recognition. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems 36, 12 (
Dec. 2017), 2017–2029.DOI: Google ScholarCross Ref - [143] . 2016. Conditional deep learning for energy-efficient and enhanced pattern recognition. In 2016 Design, Automation & Test in Europe Conference & Exhibition (DATE). 475–480.Google ScholarDigital Library
- [144] . 2017. Energy-efficient and improved image recognition with conditional deep learning. ACM Journal on Emerging Technologies in Computing Systems 13, 3 (
Feb. 2017), 33:1–33:21.DOI: Google ScholarDigital Library - [145] . 2022. DeltaCNN: End-to-End CNN inference of sparse frame differences in videos. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 12497–12506.Google ScholarCross Ref
- [146] . 2015. Big/little deep neural network for ultra low power inference. In 2015 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS). 124–132.
DOI: Google ScholarCross Ref - [147] . 2022. Dynamic Spatial Sparsification for Efficient Vision Transformers and Convolutional Neural Networks.
DOI: arxiv:2207.01580 [cs].Google ScholarCross Ref - [148] . 2017. Attention-aware deep reinforcement learning for video face recognition. In Proceedings of the IEEE International Conference on Computer Vision. 3931–3940.Google ScholarCross Ref
- [149] . 2021. DynamicViT: Efficient vision transformers with dynamic token sparsification. In Advances in Neural Information Processing Systems, Vol. 34. Curran Associates, Inc., 13937–13949.Google Scholar
- [150] . 2022. Template matching based early exit CNN for energy-efficient myocardial infarction detection on low-power wearable devices. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 6, 2 (
July 2022), 68:1–68:22.DOI: Google ScholarDigital Library - [151] . 2018. SBNet: Sparse blocks network for fast inference. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 8711–8720.Google ScholarCross Ref
- [152] . 2017. Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning.
DOI: arxiv:1711.01239 [cs].Google ScholarCross Ref - [153] . 2014. Neural decision forests for semantic image labelling. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 81–88.Google ScholarDigital Library
- [154] . 2021. Similarity-aware CNN for efficient video recognition at the edge. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (
Dec. 2021).DOI: Google ScholarDigital Library - [155] . 2022. DyFiP: Explainable AI-based dynamic filter pruning of convolutional neural networks. In Proceedings of the 2nd European Workshop on Machine Learning and Systems (EuroMLSys’22). Association for Computing Machinery, New York, NY, USA, 109–115.
DOI: Google ScholarDigital Library - [156] . 2021. AÇAI: Ascent similarity caching with approximate indexes. In 2021 33rd International Teletraffic Congress (ITC-33). 1–9.Google Scholar
- [157] . 2020. Why should we add early exits to neural networks? Cognitive Computation 12, 5 (
Sept. 2020), 954–966.DOI: Google ScholarCross Ref - [158] . 2021. ELI\(\chi\)R: Eliminating computation redundancy in CNN-based video processing. In 2021 IEEE/ACM Redefining Scalability for Diversely Heterogeneous Architectures Workshop (RSDHA). 34–44.
DOI: Google ScholarCross Ref - [159] . 2022. Confident adaptive language modeling. Advances in Neural Information Processing Systems 35 (
Dec. 2022), 17456–17472.Google Scholar - [160] . 2020. The Right Tool for the Job: Matching Model and Instance Complexities.
DOI: arxiv:2004.07453 [cs].Google ScholarCross Ref - [161] . 2018. Neural Speed Reading via Skim-RNN.
DOI: arxiv:1711.02085 [cs].Google ScholarCross Ref - [162] . 2017. Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer.
DOI: arxiv:1701.06538 [cs, stat].Google ScholarCross Ref - [163] . 2020. Fractional skipping: Towards finer-grained dynamic CNN inference. Proceedings of the AAAI Conference on Artificial Intelligence 34, 04 (
April 2020), 5700–5708.DOI: Google ScholarCross Ref - [164] . 2021. Feature-Gate Coupling for Dynamic Network Pruning.
DOI: arxiv:2111.14302 [cs].Google ScholarCross Ref - [165] . 2017. Dynamic edge-conditioned filters in convolutional neural networks on graphs. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 3693–3702.Google ScholarCross Ref
- [166] . 2020. VR-DANN: Real-time video recognition via decoder-assisted neural network acceleration. In 2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO). 698–710.
DOI: Google ScholarCross Ref - [167] . 2018. Designing adaptive neural networks for energy-constrained image classification. In 2018 IEEE/ACM International Conference on Computer-Aided Design (ICCAD). IEEE Press, San Diego, CA, USA, 1–8.
DOI: Google ScholarDigital Library - [168] . 2016. Leaving some stones unturned: Dynamic feature prioritization for activity detection in streaming video. In Computer Vision – ECCV 2016 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 783–800.
DOI: Google ScholarCross Ref - [169] . 2019. Adaptive Attention Span in Transformers.
arxiv:1905.07799 [cs, stat].Google Scholar - [170] . 2021. Dynamic network quantization for efficient video inference. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 7375–7385.Google ScholarCross Ref
- [171] . 2016. Energy-efficient adaptive classifier design for mobile systems. In Proceedings of the 2016 International Symposium on Low Power Electronics and Design (ISLPED’16). Association for Computing Machinery, New York, NY, USA, 52–57.
DOI: Google ScholarDigital Library - [172] . 2021. Efficient execution of deep neural networks on mobile devices with NPU. In Proceedings of the 20th International Conference on Information Processing in Sensor Networks (Co-Located with CPS-IoT Week 2021) (IPSN’21). Association for Computing Machinery, New York, NY, USA, 283–298.
DOI: Google ScholarDigital Library - [173] . 2022. Dynamic CNN accelerator supporting efficient filter generator with kernel enhancement and online channel pruning. In 2022 27th Asia and South Pacific Design Automation Conference (ASP-DAC). 436–441.
DOI: Google ScholarDigital Library - [174] . 2022. Arbitrary Bit-width Network: A Joint Layer-Wise Quantization and Adaptive Inference Approach.
DOI: arxiv:2204.09992 [cs].Google ScholarCross Ref - [175] . 2018. Deep progressive reinforcement learning for skeleton-based action recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 5323–5332.Google ScholarCross Ref
- [176] . 2016. Runtime configurable deep neural networks for energy-accuracy trade-off. In 2016 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS). 1–10.Google ScholarDigital Library
- [177] . 2019. Adaptive neural trees. In Proceedings of the 36th International Conference on Machine Learning. PMLR, 6166–6175.Google Scholar
- [178] . 2018. Adaptive deep learning model selection on embedded systems. ACM SIGPLAN Notices 53, 6 (
June 2018), 31–43.DOI: Google ScholarDigital Library - [179] . 2016. BranchyNet: Fast inference via early exiting from deep neural networks. In 2016 23rd International Conference on Pattern Recognition (ICPR). 2464–2469.
DOI: Google ScholarCross Ref - [180] . 2017. Distributed deep neural networks over the cloud, the edge and end devices. In 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS). 328–339.
DOI: Google ScholarCross Ref - [181] . 2021. ActionSpotter: Deep reinforcement learning framework for temporal action spotting in videos. In 2020 25th International Conference on Pattern Recognition (ICPR). 631–638.
DOI: Google ScholarCross Ref - [182] . 2018. Convolutional networks with adaptive inference graphs. In Proceedings of the European Conference on Computer Vision (ECCV). 3–18.Google ScholarDigital Library
- [183] . 2018. Shadow puppets: Cloud-level accurate {AI} inference at the speed and economy of edge. In USENIX Workshop on Hot Topics in Edge Computing (HotEdge 18).Google Scholar
- [184] . 2020. Dynamic convolutions: Exploiting spatial sparsity for faster inference. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2320–2329.Google ScholarCross Ref
- [185] . 2021. BlockCopy: High-resolution video processing with block-sparse feature propagation and online policies. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 5158–5167.Google ScholarCross Ref
- [186] . 2019. ELASTIC: Improving CNNs With dynamic scaling policies. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2258–2267.Google ScholarCross Ref
- [187] . 2018. Bandwidth-efficient live video analytics for drones via edge computing. In 2018 IEEE/ACM Symposium on Edge Computing (SEC). 159–173.
DOI: Google ScholarCross Ref - [188] . 2016. Temporal segment networks: Towards good practices for deep action recognition. In Computer Vision – ECCV 2016 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 20–36.
DOI: Google ScholarCross Ref - [189] . 2020. ECA-Net: Efficient channel attention for deep convolutional neural networks. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 11531–11539.
DOI: Google ScholarCross Ref - [190] . 2018. SkipNet: Learning dynamic routing in convolutional networks. In Proceedings of the European Conference on Computer Vision (ECCV). 409–424.Google ScholarDigital Library
- [191] . 2020. Deep mixture of experts via shallow embedding. In Proceedings of The 35th Uncertainty in Artificial Intelligence Conference. PMLR, 552–562.Google Scholar
- [192] . 2021. Adaptive focus for efficient video recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 16249–16258.Google ScholarCross Ref
- [193] . 2021. Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognition.
DOI: arxiv:2105.15075 [cs].Google ScholarCross Ref - [194] . 2020. Dual dynamic inference: Enabling more efficient, adaptive, and controllable deep inference. IEEE Journal of Selected Topics in Signal Processing 14, 4 (
May 2020), 623–633.DOI: Google ScholarCross Ref - [195] . 2019. Multi-agent reinforcement learning based frame sampling for effective untrimmed video recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 6222–6231.Google ScholarCross Ref
- [196] . 2020. Dynamic inference: A new approach toward efficient video action recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 676–677.Google ScholarCross Ref
- [197] . 2018. BlockDrop: Dynamic inference paths in residual networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 8817–8826.Google ScholarCross Ref
- [198] . 2019. LiteEval: A coarse-to-fine framework for resource efficient video recognition. In Advances in Neural Information Processing Systems, Vol. 32. Curran Associates, Inc.Google Scholar
- [199] . 2019. AdaFrame: Adaptive frame selection for fast video recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 1278–1287.Google ScholarCross Ref
- [200] . 2021. Dynamic sparsity neural networks for automatic speech recognition. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). 6014–6018.
DOI: Google ScholarCross Ref - [201] . 2022. Fully dynamic inference with deep neural networks. IEEE Transactions on Emerging Topics in Computing 10, 2 (
April 2022), 962–972.DOI: Google ScholarCross Ref - [202] . 2020. Spatially adaptive inference with stochastic feature sampling and interpolation. In Computer Vision – ECCV 2020 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 531–548.
DOI: Google ScholarDigital Library - [203] . 2020. DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference.
DOI: arxiv:2004.12993 [cs].Google ScholarCross Ref - [204] . 2021. BERxiT: Early exiting for BERT with better fine-tuning and extension to regression. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, , , and (Eds.). Association for Computational Linguistics, Online, 91–104.
DOI: Google ScholarCross Ref - [205] . 2021. Edge intelligence: Empowering intelligence to the edge of network. Proc. IEEE 109, 11 (
Nov. 2021), 1778–1837.DOI: Google ScholarCross Ref - [206] . 2015. Show, attend and tell: Neural image caption generation with visual attention. In Proceedings of the 32nd International Conference on Machine Learning. PMLR, 2048–2057.Google ScholarDigital Library
- [207] . 2020. CHA: A caching framework for home-based voice assistant systems. In 2020 IEEE/ACM Symposium on Edge Computing (SEC). 293–306.
DOI: Google ScholarCross Ref - [208] Mengwei Xu, Xuanzhe Liu, Yunxin Liu, and Felix Xiaozhu Lin. 2017. Accelerating convolutional neural networks for continuous mobile vision via cache reuse. CoRR abs/1712.01670 (2017). arXiv:1712.01670. http://arxiv.org/abs/1712.01670Google Scholar
- [209] . 2018. DeepCache: Principled cache for mobile deep vision. In Proceedings of the 24th Annual International Conference on Mobile Computing and Networking (MobiCom’18). Association for Computing Machinery, New York, NY, USA, 129–144.
DOI: Google ScholarDigital Library - [210] . 2019. Incremental training and group convolution pruning for runtime DNN performance scaling on heterogeneous embedded platforms. In 2019 ACM/IEEE 1st Workshop on Machine Learning for CAD (MLCAD). 1–6.
DOI: Google ScholarCross Ref - [211] . 2020. Optimising resource management for embedded machine learning. In 2020 Design, Automation & Test in Europe Conference & Exhibition (DATE). 1556–1561.
DOI: Google ScholarCross Ref - [212] . 2015. HD-CNN: Hierarchical deep convolutional neural networks for large scale visual recognition. In Proceedings of the IEEE International Conference on Computer Vision. 2740–2748.Google ScholarDigital Library
- [213] . 2019. cDeepArch: A compact deep neural network architecture for mobile sensing. IEEE/ACM Transactions on Networking 27, 5 (
Oct. 2019), 2043–2055.DOI: Google ScholarDigital Library - [214] . 2022. FlexPatch: Fast and accurate object detection for on-device high-resolution live video analytics. In IEEE INFOCOM 2022 - IEEE Conference on Computer Communications. 1898–1907.
DOI: Google ScholarDigital Library - [215] . 2020. Resolution adaptive networks for efficient inference. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2369–2378.Google ScholarCross Ref
- [216] . 2022. Once for all skip: Efficient adaptive deep neural networks. In 2022 Design, Automation & Test in Europe Conference & Exhibition (DATE). 568–571.
DOI: Google ScholarCross Ref - [217] . 2019. Dynamic-stride-net: Deep convolutional neural network with dynamic stride. In Optoelectronic Imaging and Multimedia Technology VI, Vol. 11187. SPIE, 42–53.
DOI: Google ScholarCross Ref - [218] . 2016. End-to-end learning of action detection from frame glimpses in videos. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2678–2687.Google ScholarCross Ref
- [219] . 2022. A-ViT: Adaptive tokens for efficient vision transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 10809–10818.Google ScholarCross Ref
- [220] . 2022. Delta Activation Layer Exploits Temporal Sparsity for Efficient Embedded Video Processing.Google ScholarCross Ref
- [221] . 2017. Learning to Skim Text.
DOI: arxiv:1704.06877 [cs].Google ScholarCross Ref - [222] . 2021. Any-precision deep neural networks. Proceedings of the AAAI Conference on Artificial Intelligence 35, 12 (
May 2021), 10763–10771.DOI: Google ScholarCross Ref - [223] . 2019. Universally slimmable networks and improved training techniques. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 1803–1811.Google ScholarCross Ref
- [224] . 2018. Slimmable Neural Networks.
DOI: arxiv:1812.08928 [cs].Google ScholarCross Ref - [225] . 2022. Fast and accurate text classification: Skimming, rereading and early stopping. (
Feb. 2022).Google Scholar - [226] . 2021. Differentiable dynamic wirings for neural networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 327–336.Google ScholarCross Ref
- [227] . 2020. S2DNAS: Transforming static CNN model for dynamic inference via neural architecture search. In Computer Vision – ECCV 2020 (Lecture Notes in Computer Science), , , , and (Eds.). Springer International Publishing, Cham, 175–192.
DOI: Google ScholarDigital Library - [228] . 2019. Boomerang: On-demand cooperative deep neural network inference for edge intelligence on the industrial internet of things. IEEE Network 33, 5 (
Sept. 2019), 96–103.DOI: Google ScholarDigital Library - [229] . 2018. FFS-VA: A fast filtering system for large-scale video analytics. In Proceedings of the 47th International Conference on Parallel Processing (ICPP 2018). Association for Computing Machinery, New York, NY, USA, 1–10.
DOI: Google ScholarDigital Library - [230] . 2020. A fast filtering mechanism to improve efficiency of large-scale video analytics. IEEE Trans. Comput. 69, 6 (
June 2020), 914–928.DOI: Google ScholarCross Ref - [231] . 2022. MVPose:Realtime multi-person pose estimation using motion vector on mobile devices. IEEE Transactions on Mobile Computing (2022), 1–1.
DOI: Google ScholarDigital Library - [232] . 2019. SCAN: A scalable neural networks framework towards compact and efficient models. In Advances in Neural Information Processing Systems, Vol. 32. Curran Associates, Inc.Google Scholar
- [233] . 2021. Dynamic runtime feature map pruning. In Pattern Recognition and Computer Vision (Lecture Notes in Computer Science), , , , , , , , and (Eds.). Springer International Publishing, Cham, 411–422.
DOI: Google ScholarDigital Library - [234] . 2021. Elf: Accelerate high-resolution mobile deep vision with content-aware parallel offloading. In Proceedings of the 27th Annual International Conference on Mobile Computing and Networking (MobiCom’21). Association for Computing Machinery, New York, NY, USA, 201–214.
DOI: Google ScholarDigital Library - [235] . 2021. Dynamic convolution pruning using pooling characteristic in convolution neural networks. In Neural Information Processing (Communications in Computer and Information Science), , , , , and (Eds.). Springer International Publishing, Cham, 558–565.
DOI: Google ScholarCross Ref - [236] . 2020. Dynamic sampling networks for efficient action recognition in videos. IEEE Transactions on Image Processing 29 (2020), 7970–7983.
DOI: Google ScholarCross Ref - [237] . 2020. BERT loses patience: Fast and robust inference with early exit. In Advances in Neural Information Processing Systems, Vol. 33. Curran Associates, Inc., 18330–18341.Google Scholar
- [238] . 2018. ECO: Efficient convolutional network for online video understanding. In Proceedings of the European Conference on Computer Vision (ECCV). 695–712.Google ScholarDigital Library
- [239] Get Your Footage. 2021. Hands Up Waving Hello Green Screen Effect | Gesture Say Hi Chroma Key in HD 4K.Google Scholar
- [240] PCV. 2022. Vehicle Detection Dataset. https://universe.roboflow.com/pcv-wndzh/vehicle-detection-bq16s. Visited on 2024-04-09.Google Scholar
Index Terms
- Adapting Neural Networks at Runtime: Current Trends in At-Runtime Optimizations for Deep Learning
Recommendations
Deep learning in neural networks
In recent years, deep artificial neural networks (including recurrent ones) have won numerous contests in pattern recognition and machine learning. This historical survey compactly summarizes relevant work, much of it from the previous millennium. ...
Continual Learning with Neural Networks: A Review
CODS-COMAD '19: Proceedings of the ACM India Joint International Conference on Data Science and Management of DataContinual learning broadly refers to the algorithms which aim to learn continuously over time across varying domains, tasks or data distributions. This is in contrast to algorithms restricted to learning a fixed number of tasks in a given domain, ...
Comments