Roboflow | How to Use Computer Vision to Control OBS Studio @Roboflow | Uploaded 1 year ago | Updated 2 hours ago
Instead of buying a stream deck or fancy remote, what if you could train an AI to help you control OBS Studio? Well, good news, by using Roboflow and the OBS Websocket, we can use gestures to control the scenes and sources of OBS. While inherently this won’t make you into a successful streamer, it might empower you to build something you otherwise thought impossible.
Read the Tutorial: blog.roboflow.com/use-computer-vision-to-control-obs
Visit the GitHub: github.com/roboflow-ai/OBS-Controller
Roboflow empowers developers to build their own computer vision applications, no matter their skillset or experience. We provide all of the tools needed to convert raw images into a custom trained computer vision model and deploy it for use in applications.
1:03 - Collect Gesture Training Data
1:16 - Start Building a Model
1:40 - Upload Your Training Data
2:20 - Annotate Your Data
3:10 - Approve Your Annotations
4:00 - Explore Augmentations
5:05 - Train Your Model
6:05 - Get Model ID and Version Number
6:20 - Get Publishable API Key
6:49 - Download Repo from GitHub
7:20 - Configure Roboflow Auth.
7:45 - OBS Scenes and Sources
9:25 - OBS Websocket Settings
9:53 - Configure WS in Roboflow.js
10:20 - Open index.html in Repo Folder
10:40 - Set + Start Virtual Camera
Instead of buying a stream deck or fancy remote, what if you could train an AI to help you control OBS Studio? Well, good news, by using Roboflow and the OBS Websocket, we can use gestures to control the scenes and sources of OBS. While inherently this won’t make you into a successful streamer, it might empower you to build something you otherwise thought impossible.
Read the Tutorial: blog.roboflow.com/use-computer-vision-to-control-obs
Visit the GitHub: github.com/roboflow-ai/OBS-Controller
Roboflow empowers developers to build their own computer vision applications, no matter their skillset or experience. We provide all of the tools needed to convert raw images into a custom trained computer vision model and deploy it for use in applications.
1:03 - Collect Gesture Training Data
1:16 - Start Building a Model
1:40 - Upload Your Training Data
2:20 - Annotate Your Data
3:10 - Approve Your Annotations
4:00 - Explore Augmentations
5:05 - Train Your Model
6:05 - Get Model ID and Version Number
6:20 - Get Publishable API Key
6:49 - Download Repo from GitHub
7:20 - Configure Roboflow Auth.
7:45 - OBS Scenes and Sources
9:25 - OBS Websocket Settings
9:53 - Configure WS in Roboflow.js
10:20 - Open index.html in Repo Folder
10:40 - Set + Start Virtual Camera
