{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/b71afe551eab40e3ab8baefd1f86a16a\" frameborder=\"0\" width=\"1316\" height=\"987\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":987,"width":1316,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":987,"thumbnail_width":1316,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/b71afe551eab40e3ab8baefd1f86a16a-1695395428220.gif","duration":181.488,"title":"Auto-Tagging Construction Equipment with Custom Trained Models","description":"In this video, I demonstrate how to use custom trained models to auto-tag telehandlers in photos for a construction equipment company. I explain the limitations of generic auto-tagging and show how to create a new image classification function using Nyckel. By manually tagging a few examples of telehandlers and non-telehandlers, we can train the model to accurately tag the rest of the photos. I also discuss the importance of continuously adding more data to improve the model's accuracy. Everything done in the video is also accessible via a simple API."}