Paul Endresen, co-founder and CEO of 3LC, demonstrates the company’s latest edge AI and vision technologies and products at the December 2024 Edge AI and Vision Alliance Forum. Specifically, Endresen demonstrates his company’s platform for dataset debugging and optimization, highlighting how the 3LC platform empowers machine learning practitioners to drastically improve their models by understanding how the model interprets data during training.
In this demo, Endresen resolves a client’s model’s false positives caused by subtle issues in labeling, showcasing how 3LC identifies problematic areas, allows seamless label adjustments within the platform, and directly integrates with training scripts for immediate feedback. By retraining after these targeted fixes, this client experienced a significant reduction in false positives and up to 75% shorter training time, thanks to the inclusion of strategically enhanced training data.
This innovative approach enables users to build more accurate and efficient AI models, even allowing for smaller architectures in some cases, without disrupting existing workflows. Discover how 3LC transforms machine learning workflows by guiding data improvement through model feedback, and learn more about 3LC’s solutions, at https://3lc.ai.
