{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deepiot-compressing-deep-neural-network","title":"DeepIoT: Compressing Deep Neural Network Structures for Sensing Systems with a Compressor-Critic Framework","arxiv_id":"1706.01215","date":"2017-06-05","proceeding":null,"authors":["Shuochao Yao","Yiran Zhao","Aston Zhang","Lu Su","Tarek Abdelzaher"],"abstract":"Recent advances in deep learning motivate the use of deep neutral networks in\nsensing applications, but their excessive resource needs on constrained\nembedded devices remain an important impediment. A recently explored solution\nspace lies in compressing (approximating or simplifying) deep neural networks\nin some manner before use on the device. We propose a new compression solution,\ncalled DeepIoT, that makes two key contributions in that space. First, unlike\ncurrent solutions geared for compressing specific types of neural networks,\nDeepIoT presents a unified approach that compresses all commonly used deep\nlearning structures for sensing applications, including fully-connected,\nconvolutional, and recurrent neural networks, as well as their combinations.\nSecond, unlike solutions that either sparsify weight matrices or assume linear\nstructure within weight matrices, DeepIoT compresses neural network structures\ninto smaller dense matrices by finding the minimum number of non-redundant\nhidden elements, such as filters and dimensions required by each layer, while\nkeeping the performance of sensing applications the same. Importantly, it does\nso using an approach that obtains a global view of parameter redundancies,\nwhich is shown to produce superior compression. We conduct experiments with\nfive different sensing-related tasks on Intel Edison devices. DeepIoT\noutperforms all compared baseline algorithms with respect to execution time and\nenergy consumption by a significant margin. It reduces the size of deep neural\nnetworks by 90% to 98.9%. It is thus able to shorten execution time by 71.4% to\n94.5%, and decrease energy consumption by 72.2% to 95.7%. These improvements\nare achieved without loss of accuracy. The results underscore the potential of\nDeepIoT for advancing the exploitation of deep neural networks on\nresource-constrained embedded devices.","url_abs":"http://arxiv.org/abs/1706.01215v3","url_pdf":"http://arxiv.org/pdf/1706.01215v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deepiot-compressing-deep-neural-network","repo_url":"https://github.com/AtenaKid/Reproducible-Deep-Learning-in-Communication","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1706.01215","atlas_url":"https://app.syntology.ai/?focus=1706.01215","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}