{"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/sigma-delta-quantized-networks","title":"Sigma Delta Quantized Networks","arxiv_id":"1611.02024","date":"2016-11-07","proceeding":null,"authors":["Peter O'Connor","Max Welling"],"abstract":"Deep neural networks can be obscenely wasteful. When processing video, a\nconvolutional network expends a fixed amount of computation for each frame with\nno regard to the similarity between neighbouring frames. As a result, it ends\nup repeatedly doing very similar computations. To put an end to such waste, we\nintroduce Sigma-Delta networks. With each new input, each layer in this network\nsends a discretized form of its change in activation to the next layer. Thus\nthe amount of computation that the network does scales with the amount of\nchange in the input and layer activations, rather than the size of the network.\nWe introduce an optimization method for converting any pre-trained deep network\ninto an optimally efficient Sigma-Delta network, and show that our algorithm,\nif run on the appropriate hardware, could cut at least an order of magnitude\nfrom the computational cost of processing video data.","url_abs":"http://arxiv.org/abs/1611.02024v2","url_pdf":"http://arxiv.org/pdf/1611.02024v2.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":"sigma-delta-quantized-networks","repo_url":"https://github.com/petered/sigma-delta","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.02024","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}