Matches in SemOpenAlex for { <https://semopenalex.org/work/W2747977736> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W2747977736 abstract "In recent years, deep neural network exhibits its powerful superiority on information discrimination in many computer vision applications. However, the capacity of deep neural network architecture is still a mystery to the researchers. Intuitively, larger capacity of neural network can always deposit more information to improve the discrimination ability of the model. But, the learnable parameter scale is not feasible to estimate the capacity of deep neural network. Due to the overfitting, directly increasing hidden nodes number and hidden layer number are already demonstrated not necessary to effectively increase the network discrimination ability. In this paper, we propose a novel measurement, named valid bits, to evaluate the capacity of deep neural networks for exploring how to quantitatively understand the deep learning and the insights behind its super performance. Specifically, our scheme to retrieve the total valid bits incorporates the skilled techniques in both training phase and inference phase. In the network training, we design decimal weight regularization and 8-bit forward quantization to obtain the integer-oriented network representations. Moreover, we develop adaptive-bitwidth and non-uniform quantization strategy in the inference phase to find the neural network capacity, total valid bits. By allowing zero bitwidth, our adaptive-bitwidth quantization can execute the model reduction and valid bits finding simultaneously. In our extensive experiments, we first demonstrate that our total valid bits is a good indicator of neural network capacity. We also analyze the impact on network capacity from the network architecture and advanced training skills, such as dropout and batch normalization." @default.
- W2747977736 created "2017-08-31" @default.
- W2747977736 creator A5028354286 @default.
- W2747977736 creator A5035679293 @default.
- W2747977736 creator A5046982010 @default.
- W2747977736 creator A5057341166 @default.
- W2747977736 date "2017-08-16" @default.
- W2747977736 modified "2023-10-01" @default.
- W2747977736 title "Deep Neural Network Capacity" @default.
- W2747977736 cites W107619411 @default.
- W2747977736 cites W1576278180 @default.
- W2747977736 cites W1686810756 @default.
- W2747977736 cites W1724438581 @default.
- W2747977736 cites W2011430131 @default.
- W2747977736 cites W2095705004 @default.
- W2747977736 cites W2097117768 @default.
- W2747977736 cites W2105767494 @default.
- W2747977736 cites W2194775991 @default.
- W2747977736 cites W2233116163 @default.
- W2747977736 cites W2919115771 @default.
- W2747977736 cites W2949117887 @default.
- W2747977736 cites W2950094539 @default.
- W2747977736 cites W2964299589 @default.
- W2747977736 hasPublicationYear "2017" @default.
- W2747977736 type Work @default.
- W2747977736 sameAs 2747977736 @default.
- W2747977736 citedByCount "0" @default.
- W2747977736 crossrefType "posted-content" @default.
- W2747977736 hasAuthorship W2747977736A5028354286 @default.
- W2747977736 hasAuthorship W2747977736A5035679293 @default.
- W2747977736 hasAuthorship W2747977736A5046982010 @default.
- W2747977736 hasAuthorship W2747977736A5057341166 @default.
- W2747977736 hasConcept C108583219 @default.
- W2747977736 hasConcept C11413529 @default.
- W2747977736 hasConcept C119857082 @default.
- W2747977736 hasConcept C136886441 @default.
- W2747977736 hasConcept C144024400 @default.
- W2747977736 hasConcept C154945302 @default.
- W2747977736 hasConcept C175202392 @default.
- W2747977736 hasConcept C19165224 @default.
- W2747977736 hasConcept C193415008 @default.
- W2747977736 hasConcept C22019652 @default.
- W2747977736 hasConcept C2776214188 @default.
- W2747977736 hasConcept C28855332 @default.
- W2747977736 hasConcept C38652104 @default.
- W2747977736 hasConcept C41008148 @default.
- W2747977736 hasConcept C50644808 @default.
- W2747977736 hasConceptScore W2747977736C108583219 @default.
- W2747977736 hasConceptScore W2747977736C11413529 @default.
- W2747977736 hasConceptScore W2747977736C119857082 @default.
- W2747977736 hasConceptScore W2747977736C136886441 @default.
- W2747977736 hasConceptScore W2747977736C144024400 @default.
- W2747977736 hasConceptScore W2747977736C154945302 @default.
- W2747977736 hasConceptScore W2747977736C175202392 @default.
- W2747977736 hasConceptScore W2747977736C19165224 @default.
- W2747977736 hasConceptScore W2747977736C193415008 @default.
- W2747977736 hasConceptScore W2747977736C22019652 @default.
- W2747977736 hasConceptScore W2747977736C2776214188 @default.
- W2747977736 hasConceptScore W2747977736C28855332 @default.
- W2747977736 hasConceptScore W2747977736C38652104 @default.
- W2747977736 hasConceptScore W2747977736C41008148 @default.
- W2747977736 hasConceptScore W2747977736C50644808 @default.
- W2747977736 hasLocation W27479777361 @default.
- W2747977736 hasOpenAccess W2747977736 @default.
- W2747977736 hasPrimaryLocation W27479777361 @default.
- W2747977736 hasRelatedWork W1931611617 @default.
- W2747977736 hasRelatedWork W1988872371 @default.
- W2747977736 hasRelatedWork W2264905057 @default.
- W2747977736 hasRelatedWork W2607376987 @default.
- W2747977736 hasRelatedWork W2739997491 @default.
- W2747977736 hasRelatedWork W2748818695 @default.
- W2747977736 hasRelatedWork W2812009592 @default.
- W2747977736 hasRelatedWork W2903143476 @default.
- W2747977736 hasRelatedWork W2912890062 @default.
- W2747977736 hasRelatedWork W2914317573 @default.
- W2747977736 hasRelatedWork W2963521187 @default.
- W2747977736 hasRelatedWork W2996923818 @default.
- W2747977736 hasRelatedWork W2999803881 @default.
- W2747977736 hasRelatedWork W3042701997 @default.
- W2747977736 hasRelatedWork W3092082681 @default.
- W2747977736 hasRelatedWork W3120868136 @default.
- W2747977736 hasRelatedWork W3164947374 @default.
- W2747977736 hasRelatedWork W3165976180 @default.
- W2747977736 hasRelatedWork W3175566627 @default.
- W2747977736 hasRelatedWork W2957984704 @default.
- W2747977736 isParatext "false" @default.
- W2747977736 isRetracted "false" @default.
- W2747977736 magId "2747977736" @default.
- W2747977736 workType "article" @default.