Algorithm-heavy work presented through a usable desktop application.
Video-analysis techniques implemented with a user-facing workflow in mind.
Computer-vision implementation paired with GUI integration and visualization.
Project Scope#
VectorView combines computer-vision implementation with interface work. That mix is useful because it forces you to care about more than algorithm correctness. The results also need to be inspectable, understandable, and exposed through an application someone can actually use.
Technical Focus#
The project computes and overlays motion fields using techniques such as EBMA and Three-Step Search, and it tracks user-defined regions of interest using feature-based and learning-based tracking approaches.
In practical terms, that meant working across:
- motion-estimation logic and evaluation metrics
- ROI selection and tracking behavior
- visualization and interaction through the UI layer
That combination is what makes the project interesting. It is not just backend analysis and not just a GUI, but the integration point between the two.
Interface Considerations#
PyQt5 was important here because the project needed a desktop workflow rather than notebook-style output. A good multimedia tool should make it easy to load input, define regions, inspect results, and compare behavior without digging through raw code every time.
That shaped the implementation. The goal was not a flashy interface, but a workflow that supported repeated testing and made the algorithms visible to the user.
What This Project Shows#
VectorView shows another side of how I work: I tend to value end-to-end usability. Even in technical or academic projects, I prefer solutions that expose the result cleanly instead of stopping at the internal implementation.