eco

We have hosted the application eco in order to run this application in our online workstations with Wine or directly.

Quick description about eco:

ECO (Efficient Convolution Operators for Tracking) is a high-performance object tracking algorithm developed by Martin Danelljan and collaborators. It is based on discriminative correlation filters and designed to handle appearance changes, occlusions, and scale variations in visual object tracking tasks. The code provides a MATLAB implementation of the ECO and ECO-HC (high-speed) variants and was one of the top performers on multiple visual tracking benchmarks.

Features:

• Implements ECO and ECO-HC tracking algorithms
• Robust to object appearance changes and occlusion
• High tracking accuracy on benchmarks like VOT and OTB
• Efficient Fourier domain computations
• Includes training and evaluation scripts
• Based on MATLAB and compatible with legacy tracking frameworks

Programming Language: MATLAB.
Categories: