# CoVA: Exploiting Compressed-Domain Analysis to Accelerate Video Analytics

arXiv: https://arxiv.org/abs/2207.00588 · HTML: https://mlsanity.com/paper/2207.00588/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 65 references that are in the map.
- First posted on arXiv 2 Jul 2022.

## Builds on (most cited first, 19 of 19)

- [Simple Online and Realtime Tracking](https://mlsanity.com/paper/1602.00763/) (2016)
- [Compressed Video Action Recognition](https://mlsanity.com/paper/1712.00636/) (2017)
- [Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor](https://mlsanity.com/paper/1504.02340/) (2015)
- [Focus: Querying Large Video Datasets with Low Latency and Low Cost](https://mlsanity.com/paper/1801.03493/) (2018)
- [BlazeIt: Optimizing Declarative Aggregation and Limit Queries for Neural Network-Based Video Analytics](https://mlsanity.com/paper/1805.01046/) (2018)
- [Fast Object Detection in Compressed Video](https://mlsanity.com/paper/1811.11057/) (2018)
- [Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training](https://mlsanity.com/paper/2108.09373/) (2021)
- [Physical Representation-based Predicate Optimization for a Visual Analytics Database](https://mlsanity.com/paper/1806.04226/) (2018)
- [Fast Video Classification via Adaptive Cascading of Deep Models](https://mlsanity.com/paper/1611.06453/) (2016)
- [Jointly Optimizing Preprocessing and Inference for DNN-based Visual Analytics](https://mlsanity.com/paper/2007.13005/) (2020)
- [VStore: A Data Store for Analytics on Large Videos](https://mlsanity.com/paper/1810.01794/) (2018)
- [THIA: Accelerating Video Analytics using Early Inference and Fine-Grained Query Planning](https://mlsanity.com/paper/2102.08481/) (2021)
- [Exploiting Temporality for Semi-Supervised Video Segmentation](https://mlsanity.com/paper/1908.11309/) (2019)
- [Video Monitoring Queries](https://mlsanity.com/paper/2002.10537/) (2020)
- [TASM: A Tile-Based Storage Manager for Video Analytics](https://mlsanity.com/paper/2006.02958/) (2020)
- [Real-time Online Multi-Object Tracking in Compressed Domain](https://mlsanity.com/paper/2204.02081/) (2022)
- [Video Analytics with Zero-streaming Cameras](https://mlsanity.com/paper/1904.12342/) (2019)
- [VSS: A Storage System for Video Analytics [Technical Report]](https://mlsanity.com/paper/2103.16604/) (2021)
- [Temporal Dynamic Appearance Modeling for Online Multi-Person Tracking](https://mlsanity.com/paper/1510.02906/) (2015)

## Built on by (most cited first, 6 of 6)

- [Edge Video Analytics: A Survey on Applications, Systems and Enabling Techniques](https://mlsanity.com/paper/2211.15751/) (2022)
- [D'ej`a Vu: Efficient Video-Language Query Engine with Learning-based Inter-Frame Computation Reuse](https://mlsanity.com/paper/2506.14107/) (2025)
- [CodecSight: Leveraging Video Codec Signals for Efficient Streaming VLM Inference](https://mlsanity.com/paper/2604.06036/) (2026)
- [DeGS: A Scalable 3DGS Architecture via Decoupled Workload Parsing and Reorganization](https://mlsanity.com/paper/2608.02099/) (2026)
- [MVTrack: Ultrafast Appearance-Free Moving Object Tracking from Compressed Bitstreams](https://mlsanity.com/paper/2608.10790/) (2026)
- [Understanding and Exploiting Diagonal Attention Sparsity in Autoregressive Image Generation](https://mlsanity.com/paper/2609.19702/) (2026)

## Abstract (arXiv)

Modern retrospective analytics systems leverage cascade architecture to mitigate bottleneck for computing deep neural networks (DNNs). However, the existing cascades suffer two limitations: (1) decoding bottleneck is either neglected or circumvented, paying significant compute and storage cost for pre-processing; and (2) the systems are specialized for temporal queries and lack spatial query support. This paper presents CoVA, a novel cascade architecture that splits the cascade computation between compressed domain and pixel domain to address the decoding bottleneck, supporting both temporal and spatial queries. CoVA cascades analysis into three major stages where the first two stages are performed in compressed domain while the last one in pixel domain. First, CoVA detects occurrences of moving objects (called blobs) over a set of compressed frames (called tracks). Then, using the track results, CoVA prudently selects a minimal set of frames to obtain the label information and only decode them to compute the full DNNs, alleviating the decoding bottleneck. Lastly, CoVA associates tracks with labels to produce the final analysis results on which users can process both temporal and spatial queries. Our experiments demonstrate that CoVA offers 4.8x throughput improvement over modern cascade systems, while imposing modest accuracy loss.