@InProceedings{8553186,
  author = {Y. Bai and S. S. Bhattacharyya and A. P. Happonen and H. Huttunen},
  booktitle = {2018 26th European Signal Processing Conference (EUSIPCO)},
  title = {Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision},
  year = {2018},
  pages = {1472-1476},
  abstract = {We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.},
  keywords = {computer vision;image classification;learning (artificial intelligence);neural nets;deep networks;prediction accuracy;training time;interesting finding;execution time;computational load;flexible control;computational graph;intermediate outputs;deep neural networks;image classification;embedded computer vision;scalable framework;elastic neural networks;Training;Estimation;Neural networks;Convolution;Detectors;Europe;Deep learning;machine learning;regularization;embedded implementations;age estimation},
  doi = {10.23919/EUSIPCO.2018.8553186},
  issn = {2076-1465},
  month = {Sep.},
  url = {https://www.eurasip.org/proceedings/eusipco/eusipco2018/papers/1570437155.pdf},
}

{"_id":"qr83iFxYRHnSCsihJ","bibbaseid":"bai-bhattacharyya-happonen-huttunen-elasticneuralnetworksascalableframeworkforembeddedcomputervision-2018","authorIDs":[],"author_short":["Bai, Y.","Bhattacharyya, S. S.","Happonen, A. P.","Huttunen, H."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["Y."],"propositions":[],"lastnames":["Bai"],"suffixes":[]},{"firstnames":["S.","S."],"propositions":[],"lastnames":["Bhattacharyya"],"suffixes":[]},{"firstnames":["A.","P."],"propositions":[],"lastnames":["Happonen"],"suffixes":[]},{"firstnames":["H."],"propositions":[],"lastnames":["Huttunen"],"suffixes":[]}],"booktitle":"2018 26th European Signal Processing Conference (EUSIPCO)","title":"Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision","year":"2018","pages":"1472-1476","abstract":"We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.","keywords":"computer vision;image classification;learning (artificial intelligence);neural nets;deep networks;prediction accuracy;training time;interesting finding;execution time;computational load;flexible control;computational graph;intermediate outputs;deep neural networks;image classification;embedded computer vision;scalable framework;elastic neural networks;Training;Estimation;Neural networks;Convolution;Detectors;Europe;Deep learning;machine learning;regularization;embedded implementations;age estimation","doi":"10.23919/EUSIPCO.2018.8553186","issn":"2076-1465","month":"Sep.","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2018/papers/1570437155.pdf","bibtex":"@InProceedings{8553186,\n author = {Y. Bai and S. S. Bhattacharyya and A. P. Happonen and H. Huttunen},\n booktitle = {2018 26th European Signal Processing Conference (EUSIPCO)},\n title = {Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision},\n year = {2018},\n pages = {1472-1476},\n abstract = {We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.},\n keywords = {computer vision;image classification;learning (artificial intelligence);neural nets;deep networks;prediction accuracy;training time;interesting finding;execution time;computational load;flexible control;computational graph;intermediate outputs;deep neural networks;image classification;embedded computer vision;scalable framework;elastic neural networks;Training;Estimation;Neural networks;Convolution;Detectors;Europe;Deep learning;machine learning;regularization;embedded implementations;age estimation},\n doi = {10.23919/EUSIPCO.2018.8553186},\n issn = {2076-1465},\n month = {Sep.},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2018/papers/1570437155.pdf},\n}\n\n","author_short":["Bai, Y.","Bhattacharyya, S. S.","Happonen, A. P.","Huttunen, H."],"key":"8553186","id":"8553186","bibbaseid":"bai-bhattacharyya-happonen-huttunen-elasticneuralnetworksascalableframeworkforembeddedcomputervision-2018","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2018/papers/1570437155.pdf"},"keyword":["computer vision;image classification;learning (artificial intelligence);neural nets;deep networks;prediction accuracy;training time;interesting finding;execution time;computational load;flexible control;computational graph;intermediate outputs;deep neural networks;image classification;embedded computer vision;scalable framework;elastic neural networks;Training;Estimation;Neural networks;Convolution;Detectors;Europe;Deep learning;machine learning;regularization;embedded implementations;age estimation"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2018url.bib","creationDate":"2021-02-13T15:38:40.298Z","downloads":0,"keywords":["computer vision;image classification;learning (artificial intelligence);neural nets;deep networks;prediction accuracy;training time;interesting finding;execution time;computational load;flexible control;computational graph;intermediate outputs;deep neural networks;image classification;embedded computer vision;scalable framework;elastic neural networks;training;estimation;neural networks;convolution;detectors;europe;deep learning;machine learning;regularization;embedded implementations;age estimation"],"search_terms":["elastic","neural","networks","scalable","framework","embedded","computer","vision","bai","bhattacharyya","happonen","huttunen"],"title":"Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision","year":2018,"dataSources":["yiZioZximP7hphDpY","iuBeKSmaES2fHcEE9"]}