← All projects
Fine-Grained Visual Recognition
Vehicle Make, Model and Year Recognition
Fine-grained vehicle recognition from large-scale image collections.
Research / engineeringOverview
A computer vision direction focused on identifying vehicle attributes such as make, model, and year. The work involves large-scale image data, data quality, and fine-grained visual classification.
Methods & topics
Highlights
- Working with large image collections for vehicle recognition.
- Considering label quality and class imbalance in dataset preparation.
- Evaluating fine-grained attribute prediction.