@InProceedings{8081219,
  author = {A. Punjabi and A. K. Katsaggelos},
  booktitle = {2017 25th European Signal Processing Conference (EUSIPCO)},
  title = {Visualization of feature evolution during convolutional neural network training},
  year = {2017},
  pages = {311-315},
  abstract = {Convolutional neural networks (CNNs) are a staple in the fields of computer vision and image processing. These networks perform visual tasks with state-of-the-art accuracy; yet, the understanding behind the success of these algorithms is still lacking. In particular, the process by which CNNs learn effective task-specific features is still unclear. This work elucidates such phenomena by applying recent deep visualization techniques during different stages of the training process. Additionally, this investigation provides visual justification to the benefits of transfer learning. The results are in line with previously discussed notions of feature specificity, and show a new facet of a particularly vexing machine learning pitfall: overfitting.},
  keywords = {computer vision;learning (artificial intelligence);neural nets;convolutional neural network training;computer vision;image processing;visual tasks;visual justification;feature specificity;CNN;deep visualization techniques;vexing machine learning pitfall;Visualization;Training;Signal processing algorithms;Neural networks;Convolution;Europe;deep learning;convolutional neural network;feature visualization;transfer learning},
  doi = {10.23919/EUSIPCO.2017.8081219},
  issn = {2076-1465},
  month = {Aug},
  url = {https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570347731.pdf},
}

{"_id":"rhG8MJQNSvMdaJnDQ","bibbaseid":"punjabi-katsaggelos-visualizationoffeatureevolutionduringconvolutionalneuralnetworktraining-2017","authorIDs":[],"author_short":["Punjabi, A.","Katsaggelos, A. K."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["A."],"propositions":[],"lastnames":["Punjabi"],"suffixes":[]},{"firstnames":["A.","K."],"propositions":[],"lastnames":["Katsaggelos"],"suffixes":[]}],"booktitle":"2017 25th European Signal Processing Conference (EUSIPCO)","title":"Visualization of feature evolution during convolutional neural network training","year":"2017","pages":"311-315","abstract":"Convolutional neural networks (CNNs) are a staple in the fields of computer vision and image processing. These networks perform visual tasks with state-of-the-art accuracy; yet, the understanding behind the success of these algorithms is still lacking. In particular, the process by which CNNs learn effective task-specific features is still unclear. This work elucidates such phenomena by applying recent deep visualization techniques during different stages of the training process. Additionally, this investigation provides visual justification to the benefits of transfer learning. The results are in line with previously discussed notions of feature specificity, and show a new facet of a particularly vexing machine learning pitfall: overfitting.","keywords":"computer vision;learning (artificial intelligence);neural nets;convolutional neural network training;computer vision;image processing;visual tasks;visual justification;feature specificity;CNN;deep visualization techniques;vexing machine learning pitfall;Visualization;Training;Signal processing algorithms;Neural networks;Convolution;Europe;deep learning;convolutional neural network;feature visualization;transfer learning","doi":"10.23919/EUSIPCO.2017.8081219","issn":"2076-1465","month":"Aug","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570347731.pdf","bibtex":"@InProceedings{8081219,\n author = {A. Punjabi and A. K. Katsaggelos},\n booktitle = {2017 25th European Signal Processing Conference (EUSIPCO)},\n title = {Visualization of feature evolution during convolutional neural network training},\n year = {2017},\n pages = {311-315},\n abstract = {Convolutional neural networks (CNNs) are a staple in the fields of computer vision and image processing. These networks perform visual tasks with state-of-the-art accuracy; yet, the understanding behind the success of these algorithms is still lacking. In particular, the process by which CNNs learn effective task-specific features is still unclear. This work elucidates such phenomena by applying recent deep visualization techniques during different stages of the training process. Additionally, this investigation provides visual justification to the benefits of transfer learning. The results are in line with previously discussed notions of feature specificity, and show a new facet of a particularly vexing machine learning pitfall: overfitting.},\n keywords = {computer vision;learning (artificial intelligence);neural nets;convolutional neural network training;computer vision;image processing;visual tasks;visual justification;feature specificity;CNN;deep visualization techniques;vexing machine learning pitfall;Visualization;Training;Signal processing algorithms;Neural networks;Convolution;Europe;deep learning;convolutional neural network;feature visualization;transfer learning},\n doi = {10.23919/EUSIPCO.2017.8081219},\n issn = {2076-1465},\n month = {Aug},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570347731.pdf},\n}\n\n","author_short":["Punjabi, A.","Katsaggelos, A. K."],"key":"8081219","id":"8081219","bibbaseid":"punjabi-katsaggelos-visualizationoffeatureevolutionduringconvolutionalneuralnetworktraining-2017","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570347731.pdf"},"keyword":["computer vision;learning (artificial intelligence);neural nets;convolutional neural network training;computer vision;image processing;visual tasks;visual justification;feature specificity;CNN;deep visualization techniques;vexing machine learning pitfall;Visualization;Training;Signal processing algorithms;Neural networks;Convolution;Europe;deep learning;convolutional neural network;feature visualization;transfer learning"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2017url.bib","creationDate":"2021-02-13T16:38:25.524Z","downloads":0,"keywords":["computer vision;learning (artificial intelligence);neural nets;convolutional neural network training;computer vision;image processing;visual tasks;visual justification;feature specificity;cnn;deep visualization techniques;vexing machine learning pitfall;visualization;training;signal processing algorithms;neural networks;convolution;europe;deep learning;convolutional neural network;feature visualization;transfer learning"],"search_terms":["visualization","feature","evolution","during","convolutional","neural","network","training","punjabi","katsaggelos"],"title":"Visualization of feature evolution during convolutional neural network training","year":2017,"dataSources":["2MNbFYjMYTD6z7ExY","uP2aT6Qs8sfZJ6s8b","ya2CyA73rpZseyrZ8","KTWAakbPXLGfYseXn","ePKPjG8C6yvpk4mEK","D8k2SxfC5dKNRFgro","7Dwzbxq93HWrJEhT6","qhF8zxmGcJfvtdeAg","fvDEHD49E2ZRwE3fb","H7crv8NWhZup4d4by","DHqokWsryttGh7pJE","vRJd4wNg9HpoZSMHD","sYxQ6pxFgA59JRhxi","w2WahSbYrbcCKBDsC","XasdXLL99y5rygCmq","3gkSihZQRfAD2KBo3","t5XMbyZbtPBo4wBGS","bEpHM2CtrwW2qE8FP","teJzFLHexaz5AQW5z"]}