Dirk Holtwick

Optimize for calls with big audiences

1

Great summary on the drawbacks of a mesh network https://bloggeek.me/webrtc-p2p-mesh/ Although technically this is true, mesh has benefits regarding resilience. A brainstorming about alternative ways besides SFU and MCU would...

Record session video

1

https://webrtc.github.io/samples/src/content/getusermedia/record/ https://github.com/webrtc/samples/blob/gh-pages/src/content/getusermedia/record/js/main.js

It would be nice to have the same settings again when coming back. It should be sufficient to have them stored globally and not per room. It will require some...

Picture-in-picture

1

https://caniuse.com/#feat=picture-in-picture https://developer.apple.com/documentation/webkitjs/adding_picture_in_picture_to_your_safari_media_controls https://googlechrome.github.io/samples/picture-in-picture/

Improve background blur

2

``` @tensorflow/tfjs-converter ^1.7.4 → ^2.0.1 @tensorflow/tfjs-core ^1.7.4 → ^2.0.1 ```

Killerfeatures ;)

3

- [x] - Blur background #17 - [ ] - Emojis #11 - [ ] - Movie like effects, like freezing frames and titles - [ ] - Audio transcriptions...

In certain scenarios like virtual classes, it is useful to have a moderator that manages when participants should talk or not. First of all it would be useful to give...

Transcription / Speech Recognition - offline only!

2

Waiting for https://developer.mozilla.org/en-US/docs/Web/API/Web_Speech_API [Tested Mozilla's Deep Speech](https://github.com/mozilla/DeepSpeech-examples). It is very promising, but not accurate enough yet. It also misses inter punctuation and is only available in English. The big Speech-to-Text...