{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,25]],"date-time":"2026-02-25T06:52:59Z","timestamp":1772002379639,"version":"3.50.1"},"reference-count":31,"publisher":"Association for Computing Machinery (ACM)","issue":"3","license":[{"start":{"date-parts":[[2018,2,9]],"date-time":"2018-02-09T00:00:00Z","timestamp":1518134400000},"content-version":"vor","delay-in-days":365,"URL":"http:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"name":"C-SPIN"},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"name":"a Semiconductor Research Corporation Program"},{"name":"MARCO and DARPA"},{"name":"one of the six centers of STARnet"},{"DOI":"10.13039\/100002418","name":"Intel Corporation","doi-asserted-by":"crossref","id":[{"id":"10.13039\/100002418","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Vannevar Bush Faculty Fellowship"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["J. Emerg. Technol. Comput. Syst."],"published-print":{"date-parts":[[2017,7,31]]},"abstract":"<jats:p>Deep-learning neural networks have proven to be very successful for a wide range of recognition tasks across modern computing platforms. However, the computational requirements associated with such deep nets can be quite high, and hence their energy-efficient implementation is of great interest. Although, traditionally, the entire network is utilized for the recognition of all inputs, we observe that the classification difficulty varies widely across inputs in real-world datasets; only a small fraction of inputs requires the full computational effort of a network, while a large majority can be classified correctly with very low effort. In this article, we propose Conditional Deep Learning (CDL), where the convolutional layer features are used to identify the variability in the difficulty of input instances and conditionally activate the deeper layers of the network. We achieve this by cascading a linear network of output neurons for each convolutional layer and monitoring the output of the linear network to decide whether classification can be terminated at the current stage or not. The proposed methodology thus enables the network to dynamically adjust the computational effort depending on the difficulty of the input data while maintaining competitive classification accuracy. The overall energy benefits for MNIST\/CIFAR10\/Tiny ImageNet datasets with state-of-the-art deep-learning architectures are 1.84 \u00d7 \/2.83 \u00d7 \/4.02 \u00d7 , respectively. We further employ the conditional approach to train deep-learning networks from scratch with integrated supervision from the additional output neurons appended at the intermediate convolutional layers. Our proposed integrated CDL training leads to an improvement in the gradient convergence behavior giving substantial error rate reduction on MNIST\/CIFAR-10, resulting in improved classification over state-of-the-art baseline networks.<\/jats:p>","DOI":"10.1145\/3007192","type":"journal-article","created":{"date-parts":[[2017,2,10]],"date-time":"2017-02-10T08:28:54Z","timestamp":1486715334000},"page":"1-21","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":40,"title":["Energy-Efficient and Improved Image Recognition with Conditional Deep Learning"],"prefix":"10.1145","volume":"13","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4167-6782","authenticated-orcid":false,"given":"Priyadarshini","family":"Panda","sequence":"first","affiliation":[{"name":"Purdue University, West Lafayette, IN"}]},{"given":"Abhronil","family":"Sengupta","sequence":"additional","affiliation":[{"name":"Purdue University, West Lafayette, IN"}]},{"given":"Kaushik","family":"Roy","sequence":"additional","affiliation":[{"name":"Purdue University, West Lafayette, IN"}]}],"member":"320","published-online":{"date-parts":[[2017,2,9]]},"reference":[{"key":"e_1_2_1_1_1","doi-asserted-by":"publisher","DOI":"10.1561\/2200000006"},{"key":"e_1_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.5555\/2354409.2354694"},{"key":"e_1_2_1_3_1","volume-title":"NIPS Workshop.","author":"Collobert Ronan","year":"2011","unstructured":"Ronan Collobert, Koray Kavukcuoglu, and Cl\u00e9ment Farabet. 2011. Torch7: A matlab-like environment for machine learning. In BigLearn, NIPS Workshop."},{"key":"e_1_2_1_4_1","doi-asserted-by":"publisher","DOI":"10.1109\/TASL.2011.2134090"},{"key":"e_1_2_1_5_1","doi-asserted-by":"publisher","unstructured":"Jeffrey Dean Greg Corrado Rajat Monga Kai Chen Matthieu Devin Mark Mao Andrew Senior Paul Tucker Ke Yang Quoc V. Le and others. 2012. Large scale distributed deep networks. In Advances in Neural Information Processing Systems. 1223--1231.","DOI":"10.5555\/2999134.2999271"},{"key":"e_1_2_1_6_1","volume-title":"Proceedings of the International Conference on Artificial Intelligence and Statistics. 249--256","author":"Glorot Xavier","year":"2010","unstructured":"Xavier Glorot and Yoshua Bengio. 2010. Understanding the difficulty of training deep feedforward neural networks. In Proceedings of the International Conference on Artificial Intelligence and Statistics. 249--256."},{"key":"e_1_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.5555\/3042817.3043084"},{"key":"e_1_2_1_8_1","volume-title":"Tiny imagenet challenge submission. CS 231N","author":"Hansen Lucas","year":"2015","unstructured":"Lucas Hansen. 2015. Tiny imagenet challenge submission. CS 231N (2015)."},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1021\/ci0342472"},{"key":"e_1_2_1_10_1","volume-title":"Deep residual learning for image recognition. arXiv preprint arXiv:1512.03385","author":"He Kaiming","year":"2015","unstructured":"Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2015a. Deep residual learning for image recognition. arXiv preprint arXiv:1512.03385 (2015)."},{"key":"e_1_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.123"},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2012.2205597"},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.1162\/neco.2006.18.7.1527"},{"key":"e_1_2_1_14_1","volume-title":"Salakhutdinov","author":"Hinton Geoffrey E.","year":"2012","unstructured":"Geoffrey E. Hinton, Nitish Srivastava, Alex Krizhevsky, Ilya Sutskever, and Ruslan R. Salakhutdinov. 2012. Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv:1207.0580 (2012)."},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2009.5459469"},{"key":"e_1_2_1_16_1","unstructured":"Alex Krizhevsky. 2009. Learning multiple layers of features from tiny images. (2009)."},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.5555\/2999134.2999257"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.5555\/1896300.1896315"},{"key":"e_1_2_1_20_1","volume-title":"Proceedings of the NIPS Workshop on Deep Learning and Unsupervised Feature Learning","volume":"2011","author":"Netzer Yuval","unstructured":"Yuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, and Andrew Y. Ng. 2011. Reading digits in natural images with unsupervised feature learning. In Proceedings of the NIPS Workshop on Deep Learning and Unsupervised Feature Learning, Vol. 2011."},{"key":"e_1_2_1_21_1","volume-title":"Prediction as a candidate for learning deep hierarchical models of data","author":"Palm Rasmus Berg","year":"2012","unstructured":"Rasmus Berg Palm. 2012. Prediction as a candidate for learning deep hierarchical models of data. Technical University of Denmark 5 (2012)."},{"key":"e_1_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.5555\/2971808.2971918"},{"key":"e_1_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.5555\/3042817.3043083"},{"key":"e_1_2_1_24_1","doi-asserted-by":"publisher","DOI":"10.1145\/2627369.2627625"},{"key":"e_1_2_1_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2014.131"},{"key":"e_1_2_1_26_1","volume-title":"Overfeat: Integrated recognition, localization and detection using convolutional networks. arXiv preprint arXiv:1312.6229","author":"Sermanet Pierre","year":"2013","unstructured":"Pierre Sermanet, David Eigen, Xiang Zhang, Micha\u00ebl Mathieu, Rob Fergus, and Yann LeCun. 2013. Overfeat: Integrated recognition, localization and detection using convolutional networks. arXiv preprint arXiv:1312.6229 (2013)."},{"key":"e_1_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"e_1_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1145\/2744769.2744904"},{"key":"e_1_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.5555\/3042817.3043055"},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","unstructured":"Jason Yosinski Jeff Clune Yoshua Bengio and Hod Lipson. 2014. How transferable are features in deep neural networks? In Advances in Neural Information Processing Systems. 3320--3328.","DOI":"10.5555\/2969033.2969197"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2010.5539957"}],"container-title":["ACM Journal on Emerging Technologies in Computing Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3007192","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3007192","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3007192","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,18]],"date-time":"2025-11-18T09:29:30Z","timestamp":1763458170000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3007192"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2017,2,9]]},"references-count":31,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2017,7,31]]}},"alternative-id":["10.1145\/3007192"],"URL":"https:\/\/doi.org\/10.1145\/3007192","relation":{},"ISSN":["1550-4832","1550-4840"],"issn-type":[{"value":"1550-4832","type":"print"},{"value":"1550-4840","type":"electronic"}],"subject":[],"published":{"date-parts":[[2017,2,9]]},"assertion":[{"value":"2016-03-01","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2016-10-01","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2017-02-09","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}