Voxel51
52 Case Studies
A Voxel51 Case Study
METU's Deep Learning and Computer Vision group faced a challenge in their medical research on inflammatory bowel disease (IBD), where doctor assessments during colonoscopy operations showed high variability. To provide more objective, real-time feedback, they turned to Voxel51 and its FiftyOne platform for assistance in managing their computer vision workflows.
Voxel51's solution involved using FiftyOne to load doctor annotations, compare them with model predictions, and visualize data distributions with embedded space tools. METU also used FiftyOne Brain to identify unique synthetic images generated by GANs. As a result, the research team gained a better understanding of their data and is now preparing to utilize the platform's video analysis capabilities for future work with video data from doctors.