Training AI with 3D Rendered Images

3D Rendered Synthetic Datasets have incredible potential. In this video Adam Kelly shares his progress with an AI (YOLACT) that is not only trained on a custom dataset, but a 3d rendered custom dataset made in Blender.

Links mentioned:
COCO Dataset Course: https://www.immersivelimit.com/course/creating-coco-datasets
CocoSynth: https://github.com/akTwelve/cocosynth
Cigarette Butts Project: https://www.immersivelimit.com/blog/training-an-ai-to-recognize-cigarette-butts
Weed Detector Project: https://www.immersivelimit.com/blog/ai-weed-detector
HDRI Haven: https://hdrihaven.com/
YOLACT: https://github.com/dbolya/yolact
YOLACT Tutorial: https://www.immersivelimit.com/tutorials/train-yolact-with-a-custom-coco-dataset

+ Follow These For Updates +
Immersive Limit Facebook: https://www.facebook.com/ImmersiveLimit/
Immersive Limit Twitter: https://twitter.com/ImmersiveLimit
Subscribe to the YouTube channel too!

+ Connect with Us! +
Adam's Twitter: https://twitter.com/aktwelve
Adam's LinkedIn: https://www.linkedin.com/in/arkelly12/
Kayla's Twitter: https://twitter.com/kayladkelly
Kayla's LinkedIn: https://www.linkedin.com/in/kayladk/

+ More Ways to Connect +
http://www.immersivelimit.com/connect

Видео Training AI with 3D Rendered Images канала Immersive Limit