{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/projectionnet-learning-efficient-on-device","title":"ProjectionNet: Learning Efficient On-Device Deep Networks Using Neural Projections","arxiv_id":"1708.00630","date":"2017-08-02","proceeding":null,"authors":["Sujith Ravi"],"abstract":"Deep neural networks have become ubiquitous for applications related to\nvisual recognition and language understanding tasks. However, it is often\nprohibitive to use typical neural networks on devices like mobile phones or\nsmart watches since the model sizes are huge and cannot fit in the limited\nmemory available on such devices. While these devices could make use of machine\nlearning models running on high-performance data centers with CPUs or GPUs,\nthis is not feasible for many applications because data can be privacy\nsensitive and inference needs to be performed directly \"on\" device.\n  We introduce a new architecture for training compact neural networks using a\njoint optimization framework. At its core lies a novel objective that jointly\ntrains using two different types of networks--a full trainer neural network\n(using existing architectures like Feed-forward NNs or LSTM RNNs) combined with\na simpler \"projection\" network that leverages random projections to transform\ninputs or intermediate representations into bits. The simpler network encodes\nlightweight and efficient-to-compute operations in bit space with a low memory\nfootprint. The two networks are trained jointly using backpropagation, where\nthe projection network learns from the full network similar to apprenticeship\nlearning. Once trained, the smaller network can be used directly for inference\nat low memory and computation cost. We demonstrate the effectiveness of the new\napproach at significantly shrinking the memory requirements of different types\nof neural networks while preserving good accuracy on visual recognition and\ntext classification tasks. We also study the question \"how many neural bits are\nrequired to solve a given task?\" using the new framework and show empirical\nresults contrasting model predictive capacity (in bits) versus accuracy on\nseveral datasets.","url_abs":"http://arxiv.org/abs/1708.00630v2","url_pdf":"http://arxiv.org/pdf/1708.00630v2.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":[],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"text-classification","task_name":"Text Classification"},{"task_slug":"text-classification-1","task_name":"text-classification"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-mnist","task":"Image Classification","dataset":"MNIST","model":"ProjectionNet","rank_in_archive_order":63,"of":81,"metrics":{"Accuracy":"95.0","Percentage error":"5.0"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1708.00630","atlas_url":"https://app.syntology.ai/?focus=1708.00630","mcp":null,"developers":"https://syntology.ai/developers","held":true,"reason":"arxiv_id_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"agent_calls":[{"tool":"get_citation_path","arguments":{"paper_1":"1708.00630"},"arguments_in":{"paper_2":"another paper's arXiv id or title"},"call":"get_citation_path(paper_1=\"1708.00630\", paper_2=\"…\")"},{"tool":"get_concepts_for_paper","arguments":{"arxiv_id":"1708.00630"},"call":"get_concepts_for_paper(arxiv_id=\"1708.00630\")"}],"by_repo":[],"by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":"the archive holds no code link for this paper","arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"908e055bf6435a6f4aaf60c523ba42144447fa0e7d8bc6e6e00c4beadabc6726","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}