Mixer: Efficiently Understanding and Retrieving Visual Content at Web-scale

Summary: Mixer: class-based features with separate production/execution layers for images and videos. Two retrieval layers enable aggregation; on Baidu, model production time halved and throughput 9.14x, with 95% precision and 97% recall for video retrieval. (summarized by gpt-5-nano on Feb 09 2026)

Paper ID: 12687
Venue: VLDB
Year: 2021
Pagerank: 5.9966613e-05
Overall Rank: 6,043 | 58.55%
DOI: 10.14778/3476311.3476371

Incoming Non-self Citations Over Time

Authors

1. An Qin (Baidu)
2. Mengbai Xiao (Shandong University)
3. Yongwei Wu (Baidu)
4. Xinjie Huang (Baidu)
5. Xiaodong Zhang (Ohio State University)

BibTeX Citation

@article{qin_vldb21,
        title = {{Mixer: Efficiently Understanding and Retrieving Visual Content at Web-scale}},
        author = {Qin, An and Xiao, Mengbai and Wu, Yongwei and Huang, Xinjie and Zhang, Xiaodong},
        journal = {PVLDB},
        series = {{VLDB} '21},
        volume = {14},
        number = {12},
        pages = {2906--2917},
        doi = {10.14778/3476311.3476371},
        url = {https://doi.org/10.14778/3476311.3476371},
        year = {2021}
}

Incoming Citations (Sorted by Pagerank)

Showing 3 of 3 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
1,631	High-Throughput Vector Similarity Search in Knowledge Graphs	2023	SIGMOD	0.00010174628
4,256	VIVA: An End-to-End System for Interactive Video Analytics	2022	CIDR	6.8018439e-05
9,915	MicroNN: An On-device Disk-resident Updatable Vector Database	2025	SIGMOD	5.1955087e-05

Outgoing Citations (Sorted by Pagerank)

Showing 11 of 11 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
284	NoScope: Optimizing Neural Network Queries over Video at Scale	2017	VLDB	0.00022370521
295	Accelerating Machine Learning Inference with Probabilistic Predicates	2018	SIGMOD	0.00022238183
406	AnalyticDB-V: A Hybrid Analytical Engine Towards Query Fusion for Structured and Unstructured Data	2020	VLDB	0.00019054532
569	BlazeIt: Optimizing Declarative Aggregation and Limit Queries for Neural Network-Based Video Analytics	2020	VLDB	0.00016348191
823	The Yin and Yang of Processing Data Warehousing Queries on GPU Devices	2013	VLDB	0.00013792901
1,607	Challenges and Opportunities in DNN-Based Video Analytics: A Demonstration of the BlazeIt Video Query Engine	2019	CIDR	0.0001022751
2,498	Mega-KV: A Case for GPUs to Maximize the Throughput of In-Memory Key-Value Stores	2015	VLDB	8.5016595e-05
2,504	Concurrent Analytical Query Processing with GPUs	2014	VLDB	8.4963369e-05
2,702	Panorama: A Data System for Unbounded Vocabulary Querying over Video	2020	VLDB	8.2342712e-05
3,756	Vaas: Video Analytics At Scale	2020	VLDB	7.1487427e-05
12,202	Fatman: Cost-saving and reliable archival storage based on volunteer resources	2014	VLDB	5.093636e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	4,618	Towards Effective Indexing for Very Large Video Sequence Database	2005	SIGMOD
2	12,371	Effective Data Co-Reduction for Multimedia Similarity Search	2011	SIGMOD
3	9,174	Qcluster: Relevance Feedback Using Adaptive Clustering for Content-Based Image Retrieval	2003	SIGMOD
4	4,459	Efficient and Cost-effective Techniques for Browsing and Indexing Large Video Databases	2000	SIGMOD
5	8,589	Optimizing Video Selection LIMIT Queries With Commonsense Knowledge	2024	VLDB
6	13,779	Studying Interaction Methodologies in Video Retrieval	2008	VLDB
7	12,214	Effective Multi-Modal Retrieval based on Stacked Auto-Encoders	2014	VLDB
8	12,419	Multiple Feature Fusion for Social Media Applications	2010	SIGMOD
9	3,365	Spatial and Temporal Constrained Ranked Retrieval over Videos	2022	VLDB
10	4,060	Inter-Media Hashing for Large-scale Retrieval from Heterogeneous Data Sources	2013	SIGMOD