Matches in SemOpenAlex for { <https://semopenalex.org/work/W3046665618> ?p ?o ?g. }
- W3046665618 endingPage "139366" @default.
- W3046665618 startingPage "139356" @default.
- W3046665618 abstract "The U-Net architecture is a state-of-the-art neural network for semantic image segmentation that is widely used in biomedical research. It is based on an encoder-decoder framework and its vanilla version shows already high performance in terms of segmentation quality. Due to its large parameter space, however, it has high computational costs on both, CPUs and GPUs. In a research setting, inference time is relevant, but not crucial for the results. However, especially in mobile, clinical applications a light and fast variant would allow deep-learning assisted, objective diagnosis at the point of care. In this work, we suggest an optimized, tiny-weight U-Net for an inexpensive hardware accelerator. We first mined the U-Net architecture to reduce computational complexity to increase runtime performance by simultaneously keeping the accuracy on a high level. Using an open, biomedical dataset for high-speed videoendoscopy (BAGLS), we show that we can dramatically reduce the parameter space and computations by over 99.8% while keeping the segmentation performance at 95% of our baseline. Using a custom upscaling routine, we further successfully deployed our optimized U-Net to an EdgeTPU hardware accelerator to gain cost-effective speed improvements on conventional computers and to showcase the applicability of EdgeTPUs for biomedical imaging processing of large images on portable devices. Combining the optimized architecture and the EdgeTPU, we gain a speedup of >79-times compared to our initial baseline while keeping high accuracy. This combination allows to provide immediate results to the clinician, especially in constrained computational environments, and an objective diagnosis at the point of care." @default.
- W3046665618 created "2020-08-07" @default.
- W3046665618 creator A5011005775 @default.
- W3046665618 creator A5088428516 @default.
- W3046665618 date "2020-01-01" @default.
- W3046665618 modified "2023-10-16" @default.
- W3046665618 title "Efficient Biomedical Image Segmentation on EdgeTPUs at Point of Care" @default.
- W3046665618 cites W1492347181 @default.
- W3046665618 cites W1546630874 @default.
- W3046665618 cites W1618567565 @default.
- W3046665618 cites W1971307783 @default.
- W3046665618 cites W1998844014 @default.
- W3046665618 cites W2003349073 @default.
- W3046665618 cites W2074342621 @default.
- W3046665618 cites W2109808161 @default.
- W3046665618 cites W2137356002 @default.
- W3046665618 cites W2138331854 @default.
- W3046665618 cites W2153004436 @default.
- W3046665618 cites W2153117647 @default.
- W3046665618 cites W2183341477 @default.
- W3046665618 cites W2531409750 @default.
- W3046665618 cites W2606722458 @default.
- W3046665618 cites W2765235648 @default.
- W3046665618 cites W2806699289 @default.
- W3046665618 cites W2884530895 @default.
- W3046665618 cites W2906403229 @default.
- W3046665618 cites W2919512338 @default.
- W3046665618 cites W2920974670 @default.
- W3046665618 cites W2924023169 @default.
- W3046665618 cites W2941207120 @default.
- W3046665618 cites W2943267175 @default.
- W3046665618 cites W2962914239 @default.
- W3046665618 cites W2963122961 @default.
- W3046665618 cites W2963136578 @default.
- W3046665618 cites W2963163009 @default.
- W3046665618 cites W2963881378 @default.
- W3046665618 cites W2964024268 @default.
- W3046665618 cites W2964054038 @default.
- W3046665618 cites W2964233199 @default.
- W3046665618 cites W2969640511 @default.
- W3046665618 cites W2981804141 @default.
- W3046665618 cites W2982083293 @default.
- W3046665618 cites W2988282777 @default.
- W3046665618 cites W2996136854 @default.
- W3046665618 cites W2996290406 @default.
- W3046665618 cites W3006598587 @default.
- W3046665618 cites W3103145119 @default.
- W3046665618 cites W3105636206 @default.
- W3046665618 doi "https://doi.org/10.1109/access.2020.3012722" @default.
- W3046665618 hasPublicationYear "2020" @default.
- W3046665618 type Work @default.
- W3046665618 sameAs 3046665618 @default.
- W3046665618 citedByCount "25" @default.
- W3046665618 countsByYear W30466656182020 @default.
- W3046665618 countsByYear W30466656182021 @default.
- W3046665618 countsByYear W30466656182022 @default.
- W3046665618 countsByYear W30466656182023 @default.
- W3046665618 crossrefType "journal-article" @default.
- W3046665618 hasAuthorship W3046665618A5011005775 @default.
- W3046665618 hasAuthorship W3046665618A5088428516 @default.
- W3046665618 hasBestOaLocation W30466656181 @default.
- W3046665618 hasConcept C108583219 @default.
- W3046665618 hasConcept C111919701 @default.
- W3046665618 hasConcept C113775141 @default.
- W3046665618 hasConcept C11413529 @default.
- W3046665618 hasConcept C118505674 @default.
- W3046665618 hasConcept C124504099 @default.
- W3046665618 hasConcept C154945302 @default.
- W3046665618 hasConcept C173608175 @default.
- W3046665618 hasConcept C2776214188 @default.
- W3046665618 hasConcept C31972630 @default.
- W3046665618 hasConcept C41008148 @default.
- W3046665618 hasConcept C45374587 @default.
- W3046665618 hasConcept C68339613 @default.
- W3046665618 hasConcept C79403827 @default.
- W3046665618 hasConcept C89600930 @default.
- W3046665618 hasConceptScore W3046665618C108583219 @default.
- W3046665618 hasConceptScore W3046665618C111919701 @default.
- W3046665618 hasConceptScore W3046665618C113775141 @default.
- W3046665618 hasConceptScore W3046665618C11413529 @default.
- W3046665618 hasConceptScore W3046665618C118505674 @default.
- W3046665618 hasConceptScore W3046665618C124504099 @default.
- W3046665618 hasConceptScore W3046665618C154945302 @default.
- W3046665618 hasConceptScore W3046665618C173608175 @default.
- W3046665618 hasConceptScore W3046665618C2776214188 @default.
- W3046665618 hasConceptScore W3046665618C31972630 @default.
- W3046665618 hasConceptScore W3046665618C41008148 @default.
- W3046665618 hasConceptScore W3046665618C45374587 @default.
- W3046665618 hasConceptScore W3046665618C68339613 @default.
- W3046665618 hasConceptScore W3046665618C79403827 @default.
- W3046665618 hasConceptScore W3046665618C89600930 @default.
- W3046665618 hasFunder F4320320879 @default.
- W3046665618 hasFunder F4320323803 @default.
- W3046665618 hasLocation W30466656181 @default.
- W3046665618 hasLocation W30466656182 @default.
- W3046665618 hasLocation W30466656183 @default.
- W3046665618 hasOpenAccess W3046665618 @default.
- W3046665618 hasPrimaryLocation W30466656181 @default.