VSS: A Storage System for Video Analytics

We present a new video storage system (VSS) designed to decouple high-level video operations from the low-level details required to store and efficiently retrieve video data. VSS is designed to be the storage subsystem of a video data management system (VDBMS) and is responsible for: (1) transparently and automatically arranging the data on disk in an efficient, granular format; (2) caching frequently-retrieved regions in the most useful formats; and (3) eliminating redundancies found in videos captured from multiple cameras with overlapping fields of view. Our results suggest that VSS can improve VDBMS read performance by up to 54%, reduce storage costs by up to 45%, and enable developers to focus on application logic rather than video storage and retrieval.

[1]  Sungwoo Hong,et al.  Storage technique for real-time streaming of layered video , 2009, Multimedia Systems.

[2]  Alvin Cheung,et al.  Visual Road: A Video Data Management Benchmark , 2019, SIGMOD Conference.

[3]  Jason Flinn,et al.  quFiles: The right file at the right time , 2010, TOS.

[4]  Nick Koudas,et al.  SVQ: Streaming Video Queries , 2019, SIGMOD Conference.

[5]  Thomas F. Wenisch,et al.  Physical Representation-Based Predicate Optimization for a Visual Analytics Database , 2018, 2019 IEEE 35th International Conference on Data Engineering (ICDE).

[6]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[7]  Paramvir Bahl,et al.  Focus: Querying Large Video Datasets with Low Latency and Low Cost , 2018, OSDI.

[8]  Andrew A. Chien,et al.  Networked Cameras Are the New Big Data Clusters , 2019, HotEdgeVideo@MOBICOM.

[9]  Yuhao Zhang,et al.  Panorama: A Data System for Unbounded Vocabulary Querying over Video , 2019, Proc. VLDB Endow..

[10]  Houqiang Li,et al.  Overview of the multiview high efficiency video coding (MV-HEVC) standard , 2015, 2015 IEEE International Conference on Image Processing (ICIP).

[11]  Paramvir Bahl,et al.  Live Video Analytics at Scale with Approximation and Delay-Tolerance , 2017, NSDI.

[12]  Alvin Cheung,et al.  VisualWorldDB: A DBMS for the Visual World , 2020, CIDR.

[13]  Paolo Bestagini,et al.  Near-duplicate detection and alignment for multi-view videos , 2015, 2015 IEEE International Conference on Image Processing (ICIP).

[14]  Peter Bailis,et al.  BlazeIt: Optimizing Declarative Aggregation and Limit Queries for Neural Network-Based Video Analytics , 2018, Proc. VLDB Endow..

[15]  Aakanksha Chowdhery,et al.  Accelerating Machine Learning Inference with Probabilistic Predicates , 2018, SIGMOD Conference.

[16]  Alvin Cheung,et al.  LightDB: A DBMS for Virtual Reality Video , 2018, Proc. VLDB Endow..

[17]  Christina R. Strong,et al.  Addressing the Dark Side of Vision Research: Storage , 2017, HotStorage.

[18]  Ion Stoica,et al.  Chameleon: scalable adaptation of video analytics , 2018, SIGCOMM.

[19]  Tian Zhang,et al.  BIRCH: an efficient data clustering method for very large databases , 1996, SIGMOD '96.

[20]  Hong-Yuan Mark Liao,et al.  YOLOv4: Optimal Speed and Accuracy of Object Detection , 2020, ArXiv.

[21]  Pedro A. Amado Assunção,et al.  A pixel-based complexity model to estimate energy consumption in video decoders , 2018, 2018 IEEE International Conference on Consumer Electronics (ICCE).

[22]  Alvin Cheung,et al.  TASM: A Tile-Based Storage Manager for Video Analytics , 2020, ArXiv.

[23]  Jingjing Wang,et al.  Deluceva: Delta-Based Neural Network Inference for Fast Video Analytics , 2020, SSDBM.

[24]  Tim Kraska,et al.  MIRIS: Fast Object Track Queries in Video , 2020, SIGMOD Conference.

[25]  Parthasarathy Ranganathan,et al.  vbench: Benchmarking Video Transcoding in the Cloud , 2018, ASPLOS.

[26]  Felix Xiaozhu Lin,et al.  VStore: A Data Store for Analytics on Large Videos , 2018, EuroSys.

[27]  Paolo Bestagini,et al.  Detection and Synchronization of Video Sequences for Event Reconstruction , 2019, 2019 IEEE International Conference on Image Processing (ICIP).

[28]  Djemel Ziou,et al.  Image Quality Metrics: PSNR vs. SSIM , 2010, 2010 20th International Conference on Pattern Recognition.

[29]  Paul Newman,et al.  1 year, 1000 km: The Oxford RobotCar dataset , 2017, Int. J. Robotics Res..

[30]  Alvin Cheung,et al.  VSS: A Storage System for Video Analytics , 2021, SIGMOD Conference.

[31]  Christina R. Strong,et al.  VDMS: Efficient Big-Visual-Data Access for Machine Learning Workloads , 2018, 1810.11832.

[32]  Ali Farhadi,et al.  Video Relationship Reasoning Using Gated Spatio-Temporal Energy Graph , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[33]  Pat Hanrahan,et al.  Scanner: Efficient Video Analysis at Scale , 2018, ACM Trans. Graph..

[34]  Sanjeev Kumar,et al.  Finding a Needle in Haystack: Facebook's Photo Storage , 2010, OSDI.

[35]  Peter Bailis,et al.  Jointly Optimizing Preprocessing and Inference for DNN-based Visual Analytics , 2020, Proc. VLDB Endow..

[36]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[37]  Peter Bailis,et al.  NoScope: Optimizing Deep CNN-Based Queries over Video Streams at Scale , 2017, Proc. VLDB Endow..

[38]  Alon Y. Halevy,et al.  Answering queries using views: A survey , 2001, The VLDB Journal.

[39]  Aakanksha Chowdhery,et al.  Optasia: A Relational Platform for Efficient Large-Scale Video Analytics , 2016, SoCC.

[40]  Mor Harchol-Balter,et al.  Practical Bounds on Optimal Caching with Variable Object Sizes , 2017, SIGMETRICS.

[41]  Nikolaj Bjørner,et al.  Z3: An Efficient SMT Solver , 2008, TACAS.

[42]  VStore , 2019, Proceedings of the Fourteenth EuroSys Conference 2019.

[43]  Alvin Cheung,et al.  Perceptual Compression for Video Storage and Processing Systems , 2019, SoCC.

[44]  Bill J. Yates.  Body Worn Cameras: Making Them Mandatory , 2018 .

[45]  Martha A. Kim,et al.  vbench: Benchmarking Video Transcoding in the Cloud , 2018, ASPLOS.

[46]  Calton Pu,et al.  ODIN , 2020, Proc. VLDB Endow..