DC Field | Value | Language |
---|---|---|
dc.contributor.author | Im, Dongseok | ko |
dc.contributor.author | Park, Gwangtae | ko |
dc.contributor.author | Ryu, Junha | ko |
dc.contributor.author | Li, Zhiyong | ko |
dc.contributor.author | Kang, Sanghoon | ko |
dc.contributor.author | Han, Donghyeon | ko |
dc.contributor.author | Lee, Jinsu | ko |
dc.contributor.author | Park, Wonhoon | ko |
dc.contributor.author | Kwon, Hankyul | ko |
dc.contributor.author | Yoo, Hoi-Jun | ko |
dc.date.accessioned | 2023-01-09T03:00:18Z | - |
dc.date.available | 2023-01-09T03:00:18Z | - |
dc.date.created | 2022-11-28 | - |
dc.date.issued | 2023-01 | - |
dc.identifier.citation | IEEE JOURNAL OF SOLID-STATE CIRCUITS, v.58, no.1, pp.177 - 188 | - |
dc.identifier.issn | 0018-9200 | - |
dc.identifier.uri | http://hdl.handle.net/10203/304141 | - |
dc.description.abstract | 3-D red, green, blue, and depth (RGB-D) and 3-D perception are essential information for 3-D applications such as autonomous driving and augmented reality (AR)/virtual reality (VR) systems. However, battery-and resource-limited mobile devices face difficulties in obtaining dense RGB-D data and 3-D perception information in low-power (LP) and real-time. Specifically, an RGB-D sensor is used to acquire 3-D RGB-D data, but it consumes high power and produces sparse depth data. Moreover, preprocessing for RGB-D data requires a long execution time. Previous 3-D perception accelerators also have limited reconfigurability, making them incapable of executing diverse 3-D perception tasks. In this article, an LP and real-time depth signal processing system-on-chip (SoC), depth signal processing unit (DSPU), is presented. The DSPU produces accurate dense RGB-D data using a convolutional neural network (CNN)-based monocular depth estimation (MDE) and a sensor fusion with an LP ToF sensor. Then, the DSPU performs 3-D perception inferring a point cloud-based neural network (PNN). The DSPU executes the depth signal processing system with the following features: 1) a unified point processing unit (UPPU) with a flexible window based-search algorithm for simplifying the complexity of point processing algorithms and saving the arithmetic units and buffers; 2) a unified matrix processing unit (UMPU) with bit-slice-level sparsity exploitation to accelerate various matrix processing algorithms; 3) a band matrix encoder and decoder to decrease the data transactions in the conjugate-gradient (C-Grad) method; and 4) a point feature (PF) reuse method with a pipelined architecture for low-latency and LP PNN inference. Finally, the DSPU achieves real-time implementation with 281.6 mW of the end-to-end 3-D B-box extraction system. | - |
dc.language | English | - |
dc.publisher | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC | - |
dc.title | DSPU: An Efficient Deep Learning-Based Dense RGB-D Data Acquisition With Sensor Fusion and 3-D Perception SoC | - |
dc.type | Article | - |
dc.identifier.wosid | 000881959300001 | - |
dc.identifier.scopusid | 2-s2.0-85141643186 | - |
dc.type.rims | ART | - |
dc.citation.volume | 58 | - |
dc.citation.issue | 1 | - |
dc.citation.beginningpage | 177 | - |
dc.citation.endingpage | 188 | - |
dc.citation.publicationname | IEEE JOURNAL OF SOLID-STATE CIRCUITS | - |
dc.identifier.doi | 10.1109/JSSC.2022.3218278 | - |
dc.contributor.localauthor | Yoo, Hoi-Jun | - |
dc.contributor.nonIdAuthor | Lee, Jinsu | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | Signal processing algorithms | - |
dc.subject.keywordAuthor | Sensor fusion | - |
dc.subject.keywordAuthor | Signal processing | - |
dc.subject.keywordAuthor | Sparse matrices | - |
dc.subject.keywordAuthor | Real-time systems | - |
dc.subject.keywordAuthor | Decoding | - |
dc.subject.keywordAuthor | Convolution | - |
dc.subject.keywordAuthor | 3-D perception | - |
dc.subject.keywordAuthor | conjugated-gradient (C-Grad) | - |
dc.subject.keywordAuthor | convolutional neural network (CNN) | - |
dc.subject.keywordAuthor | deep neural network | - |
dc.subject.keywordAuthor | depth estimation | - |
dc.subject.keywordAuthor | point cloud-based neural network (PNN) | - |
dc.subject.keywordAuthor | red | - |
dc.subject.keywordAuthor | green | - |
dc.subject.keywordAuthor | blue | - |
dc.subject.keywordAuthor | and depth (RGB-D) data | - |
dc.subject.keywordAuthor | sensor fusion | - |
dc.subject.keywordAuthor | system-on-chip (SoC) | - |
dc.subject.keywordPlus | ACCURATE | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.