EVA is a visual data management system (think MySQL for videos). It supports a declarative language similar to SQL and a wide range of commonly used computer vision models
- Build super apps over your video data by leveraging EVA’s simple SQL-like interface.
- Trying to find a tradeoff between throughput and accuracy on your video models? EVA can give you both:
- EVA improves throughput by introducing sampling, filtering, and caching techniques.
- EVA improves accuracy by introducing state-of-the-art model specialization and selection algorithms.
- EVA offers seamless integration into your existing model workflows with the power of EVA UDF’s. (Need to change wording from UDF to something else).
EVA consists of four core components:
- EVAQL Query Parser
- Query Optimizer
- Query Execution Engine (Filters + Deep Learning Models)
- Distributed Storage Engine
§2. Google Colab Version
The Google Colab Version for EVA enables user to easily make use of our database system without installing it in their local machine
EVAQL Query Parser
Query Execution Engine (Filters + Deep Learning Models)
Distributed Storage Engine
- The Design and Implementation of a Non-Volatile Memory Database Management System
- FiGO: Fine-Grained Query Optimization in Video Analytics
Jiashen Cao, Ramyad Hadidi, Joy Arulraj, and Hyesoon Kim
- EVA: A Symbolic Approach to Accelerating Exploratory Video Analytics with Materialized Views
Zhuangdi Xu, Gaurav Kakkar, Joy Arulraj, and Umakishore Ramachandran
- Automatic Detection of Performance Bugs in Database Systems using Equivalent Queries
Xinyu Liu, Qi Zhou, Joy Arulraj, and Alessandro Orso
- Sia: Optimizing Queries using Learned Predicates
Qi Zhou, Joy Arulraj, Shamkant Navathe, William Harris, and Qiupeng Wu
- Spitfire: A Three-Tier Buffer Manager for Volatile and Non-Volatile Memory
Xinjing Zhou, Joy Arulraj, Andy Pavlo, and David Cohen
- ODIN: Automated Drift Detection and Recovery in Video Analytics
Abhijit Suprem, Joy Arulraj, Calton Pu, and Joao Ferreira
To file a bug or request a feature, please file a GitHub issue. Pull requests are welcome.
See the people page for the full listing of contributors.