Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891465615> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W2891465615 abstract "Deploying deep models on embedded devices has been a challenging problem since the great success of deep learning based networks. Fixed-point networks, which represent their data with low bits fixed-point and thus give remarkable savings on memory usage, are generally preferred. Even though current fixed-point networks employ relative low bits (e.g. 8-bits), the memory saving is far from enough for the embedded devices. On the other hand, quantization deep networks, for example XNOR-Net and HWGQ-Net, quantize the data into 1 or 2 bits resulting in more significant memory savings but still contain lots of floating-point data. In this paper, we propose a fixed-point network for embedded vision tasks through converting the floating-point data in a quantization network into fixed-point. Furthermore, to overcome the data loss caused by the conversion, we propose to compose floating-point data operations across multiple layers (e.g. convolution, batch normalization and quantization layers) and convert them into fixed-point. We name the fixed-point network obtained through such integrated conversion as Integrated Fixed-point Quantization Networks (IFQ-Net). We demonstrate that our IFQ-Net gives 2.16× and 18× more savings on model size and runtime feature map memory respectively with similar accuracy on ImageNet. Furthermore, based on YOLOv2, we design IFQ-Tinier-YOLO face detector which is a fixed-point network with 256× reduction in model size (246k Bytes) than Tiny-YOLO. We illustrate the promising performance of our face detector in terms of detection rate on Face Detection Data Set and Bencmark (FDDB) and qualitative results of detecting small faces of Wider Face dataset." @default.
- W2891465615 created "2018-09-27" @default.
- W2891465615 creator A5000695768 @default.
- W2891465615 creator A5002027079 @default.
- W2891465615 creator A5013433372 @default.
- W2891465615 creator A5046355512 @default.
- W2891465615 creator A5049217397 @default.
- W2891465615 creator A5086129126 @default.
- W2891465615 date "2018-06-01" @default.
- W2891465615 modified "2023-09-22" @default.
- W2891465615 title "IFQ-Net: Integrated Fixed-Point Quantization Networks for Embedded Vision" @default.
- W2891465615 cites W182571476 @default.
- W2891465615 cites W1999085092 @default.
- W2891465615 cites W2286365479 @default.
- W2891465615 cites W2300242332 @default.
- W2891465615 cites W2337344472 @default.
- W2891465615 cites W2473118437 @default.
- W2891465615 cites W2570343428 @default.
- W2891465615 cites W2586654419 @default.
- W2891465615 cites W2778795994 @default.
- W2891465615 cites W2963114950 @default.
- W2891465615 cites W2963150697 @default.
- W2891465615 cites W2963374099 @default.
- W2891465615 cites W2963566548 @default.
- W2891465615 cites W2964200805 @default.
- W2891465615 cites W3106250896 @default.
- W2891465615 cites W639708223 @default.
- W2891465615 doi "https://doi.org/10.1109/cvprw.2018.00103" @default.
- W2891465615 hasPublicationYear "2018" @default.
- W2891465615 type Work @default.
- W2891465615 sameAs 2891465615 @default.
- W2891465615 citedByCount "12" @default.
- W2891465615 countsByYear W28914656152019 @default.
- W2891465615 countsByYear W28914656152020 @default.
- W2891465615 countsByYear W28914656152021 @default.
- W2891465615 crossrefType "proceedings-article" @default.
- W2891465615 hasAuthorship W2891465615A5000695768 @default.
- W2891465615 hasAuthorship W2891465615A5002027079 @default.
- W2891465615 hasAuthorship W2891465615A5013433372 @default.
- W2891465615 hasAuthorship W2891465615A5046355512 @default.
- W2891465615 hasAuthorship W2891465615A5049217397 @default.
- W2891465615 hasAuthorship W2891465615A5086129126 @default.
- W2891465615 hasBestOaLocation W28914656152 @default.
- W2891465615 hasConcept C108583219 @default.
- W2891465615 hasConcept C11413529 @default.
- W2891465615 hasConcept C134306372 @default.
- W2891465615 hasConcept C154945302 @default.
- W2891465615 hasConcept C28855332 @default.
- W2891465615 hasConcept C33923547 @default.
- W2891465615 hasConcept C41008148 @default.
- W2891465615 hasConcept C43364308 @default.
- W2891465615 hasConcept C61445026 @default.
- W2891465615 hasConcept C76155785 @default.
- W2891465615 hasConcept C79403827 @default.
- W2891465615 hasConcept C9390403 @default.
- W2891465615 hasConcept C94915269 @default.
- W2891465615 hasConceptScore W2891465615C108583219 @default.
- W2891465615 hasConceptScore W2891465615C11413529 @default.
- W2891465615 hasConceptScore W2891465615C134306372 @default.
- W2891465615 hasConceptScore W2891465615C154945302 @default.
- W2891465615 hasConceptScore W2891465615C28855332 @default.
- W2891465615 hasConceptScore W2891465615C33923547 @default.
- W2891465615 hasConceptScore W2891465615C41008148 @default.
- W2891465615 hasConceptScore W2891465615C43364308 @default.
- W2891465615 hasConceptScore W2891465615C61445026 @default.
- W2891465615 hasConceptScore W2891465615C76155785 @default.
- W2891465615 hasConceptScore W2891465615C79403827 @default.
- W2891465615 hasConceptScore W2891465615C9390403 @default.
- W2891465615 hasConceptScore W2891465615C94915269 @default.
- W2891465615 hasLocation W28914656151 @default.
- W2891465615 hasLocation W28914656152 @default.
- W2891465615 hasOpenAccess W2891465615 @default.
- W2891465615 hasPrimaryLocation W28914656151 @default.
- W2891465615 hasRelatedWork W2078345533 @default.
- W2891465615 hasRelatedWork W2223186295 @default.
- W2891465615 hasRelatedWork W2612445135 @default.
- W2891465615 hasRelatedWork W2752698706 @default.
- W2891465615 hasRelatedWork W2784372305 @default.
- W2891465615 hasRelatedWork W2786771851 @default.
- W2891465615 hasRelatedWork W2893550440 @default.
- W2891465615 hasRelatedWork W2918006316 @default.
- W2891465615 hasRelatedWork W2960955628 @default.
- W2891465615 hasRelatedWork W2963983756 @default.
- W2891465615 hasRelatedWork W2964299589 @default.
- W2891465615 hasRelatedWork W2968095833 @default.
- W2891465615 hasRelatedWork W2991110335 @default.
- W2891465615 hasRelatedWork W3117528324 @default.
- W2891465615 hasRelatedWork W3136839710 @default.
- W2891465615 hasRelatedWork W3140099337 @default.
- W2891465615 hasRelatedWork W3160276637 @default.
- W2891465615 hasRelatedWork W3165412284 @default.
- W2891465615 hasRelatedWork W3180266648 @default.
- W2891465615 hasRelatedWork W3211659500 @default.
- W2891465615 isParatext "false" @default.
- W2891465615 isRetracted "false" @default.
- W2891465615 magId "2891465615" @default.
- W2891465615 workType "article" @default.