CCTV | C Compile-Time eValuator for scheme | Interpreter library

I would break your useEffect to two since they handle different parts of your code, one to dispatch a fetch and the other to massage the data when received.

You have to choose a balance between encoding speed and encoding efficiency.

1. Choose the slowest -preset you have patience for.
2. Choose the highest -crf value that provides an acceptable quality.

See FFmpeg Wiki: H.264 for more info.

If libx264 does not make a small enough file try libx265, but it takes longer to encode.

See FFmpeg Wiki: HEVC/H.265 for more info.

If you have the proper hardware, then you can use NVENC, QuickSync, or some other implementation.

Encoding will be fast, but it will not match the quality per bit provided by libx264 or libx265.

See FFmpeg Wiki: Hardware for more info.

Available RAM is not the only thing that's influenced by the Memory setting, which is a bit counterintuitive at first. The memory setting in a Lambda function influences:

• The amount of compute (vCPUs) that's available to the function
• The amount of network throughput that's available
• The amount of system memory / RAM that's available

All of these scale based on the memory. You can think of it like this: If you provision 128MB RAM, you also get about 1/8 of a vCPU, if you go to 256 MB, you get a quarter of a vCPU. These are not exact numbers, but it's a useful mental model. Somewhere between 1024 and 1280 MB you get a full vCPU and afterwards a second vCPU is added.

Your workload seems CPU intensive (since there's not GPU to offload it to), so I'd try to increase the Memory to give Lambda more compute power to see how it behaves.

In each frame you are adding (turnSpeed * Time.deltaTime) to the current angle. In the likely case that it passes 300 or 0 (For example 298.5f + 2.5f will pass 300) the camera will start to turn clockwise and counter-clockwise each frame. (298.5f, 301.0f, 298.5f, 301.0f and so on). You should be using mathematical functions that we are sure that will not exceed some values. Sin() function is a good example since we know that it will always be in the range of -1 and 1. Something like this should give you the idea:

It looks like you're using another version of three.js and referencing it globally, which is not delivered by Forge Viewer itself, and it will conflict with Forge Viewer one. As Petr mentioned here (upgrade three js version in autodesk forge viewer), Forge Viewer uses self-maintained three.js r71 removed most of the built-in three.js functions.

If you want to use missing features from Forge Viewer's three.js libraries, I would advise you to use some modern development tools such as bundlers (Webpack is the most popular one). You can leverage ES6 modules and import to your app only the strictly needed dependencies. The threejs-full-es6 package lets you import Three.js features to your app independently, so in that case, we can import THREEE.ObjectLoader, and all other required dependencies will be taken care of by the package implementation.

See Forge community blog here: https://forge.autodesk.com/blog/how-add-newest-threejs-features-forge-viewer

images in this website are not present in the initial html file but loaded dynamically through execution of javascript and beautifulsoup/urllib are not going to execute them for you.

to crawl dynamic websites you should use a sort of headless browser like selenium which has python library to communicate with it. these browsers are just like normal browsers but with one difference; they are controlled by your code instead of the user.

a better alternative to selenium would be puppeteer but i used it in node.js and i'm not sure of it's python binding quality.

I think that it could be help you do save the images in your Pc:

So because the camera is projecting a 3D world down to 2D and then you want to get it from another 2D perspective there can be errors. For instance the center point of the rectangle of the person could either be ~2.7ft off the ground where they are standing OR could be the floor if the person wasn't there. Those would be 2 very different places on your birds eye view map.

However, since this is for tracking people, you could make the assumption that everyone is approximately the same height and assume the center of the rectangle is always approximately 2.7ft off the ground. If you make this assumption then the problem is more tractable.

With that assumption, what you could do is have a calibration phase. So you could have the person stand at the end of the hallway and see what coordinates that is on the camera and what coordinates on the map. Then have them walk up in front of the camera and see the coordinates on the camera and coordinates on the map. With these two points you could do a linear interpolation to be able to figure out where in the hallway someone is based on the camera. You would need to do this calibration for each camera that you have but it should give fairly accurate results.

Let (x1, y1) be the camera coordinates at end of hall and (X1, Y1) map coordinates at end of hall. Then let (x2, y2) be camera coordinates when close to camera and (X2, Y2) map coordinates close to camera. So then find a linear A such that

A(x1, y1) = (X1, Y1) and A(x2, y2) = (X2, Y2)

You can solve this as a matrix equation (not sure how to type math in here)