Webcam | Make an android phone a WiFi webcam | Camera library

You need to set the size of the actual canvas, like this:

If you connect the button's clicked signal to the worker's start slot, Qt will automatically detect that it's a cross-thread connection. When the signal is eventually emitted, it will be queued in the receiving thread's event-queue, which ensures the slot will be called within the worker thread.

However, if you connect the button's clicked signal to the startCapture slot, there's no cross-thread connection, because the slot belongs to MyWidget (which lives in the main thread). When the signal is emitted this time, the slot tries to create the timer from within the main thread, which is not supported. Timers must always be started within the thread that creates them (otherwise Qt will print a message like "QBasicTimer::start: Timers cannot be started from another thread").

A better approach is to connect the started and finished signals of the thread to some start and stop slots in the worker, and then call the thread's start and quit methods to control the worker. Here's a demo based on your script, which shows how to implement that:

I figured it out. You better not hold anything in your mouth, because this is hilarious!

Basically being all out of options, I locked on to the last discernible difference between this deployment and the ones that worked: The filesize of the jar being deployed. The one that failed was by far the smallest. So I bloated it up by some 60% to make it comparable to everything else... and that fixed it!

This sounds preposterous. Here's my hypothesis on what's going on: If the upload takes too little time, the lambda somehow needs longer to change its state. I'm not sure why that would be, you'd expect the state to change when things are done, not to take longer if things are done faster, right? Maybe there's a minimum time for the state to remain? I wouldn't know. There's one thing to support this hypothesis, though: The deployment from my local computer always worked. That upload would naturally take longer than jenkins needs from inside the aws vpc. So this hypothesis, as ludicrous as it sounds, fits all the facts that I have on hand.

Maybe somebody with a better understanding of the lambda-internal mechanisms can add a comment to this explaining how this can happen...

Is there a way to push my local Websocket to a live environment?

It's not straightforward because you need more than vanilla webrtc (which is peer-to-peer). What you want is an SFU. Take a look at mediasoup.

To realize why this is needed think about how the webrtc connection is established in your current app. It's a negotiation between two parties (facilitated by a signaling server). In order to turn this into a multi-cast setup you will need a proxy of sorts that then establishes separate peer-to-peer connections to all senders and receivers.

Maybe try changing the shape of pose_landmarks from (33, 3) to (1, 99) after your assertion and before you make a prediction:

I fixed that by avenc_aac instead of fdkaacenc. Not to mention that you should use audioconvert before avenc_aac.

The part of command I was trying to convert to aac:

rtpopusdepay ! opusdec ! audioconvert ! avenc_aac.

More: avenc_aac is part of gstreamer ffmpeg plugins which exists in linux, mac and windows. fdkaacenc is part of gstreamer bad plugins that may not be the same in all systems. That's is why using avenc_aac is more approperiate. So it can be executed in both mac and linux.

The error message was triggered due to the offer containing H264 codecs whilst the Android Client was not anticipating H264 and was not setup to encode and/or decode this particular hardware encoded stream.

The fix was to ensure that the connection factory was setup as such:

tl;dr: This is how. c# doesn't need to be aware of the async and it should work.

I just made a little test using

Assets/Plugins/mylib.jslib

Short Implementation of Christoph's great solution in the comments.

Step 1: Use the paint tool to extract the coordinates of the black box region.

you can see the x,y coordinates at the bottom of your image.

Step 2:

a. resize every frame of the video to fit the black region size. width = x1 - x0 and height = y1 - y0.

b. replace the black region pixels with the resized frames.

code: