git-dive VS mediapipe

> Some advices for a first timer?

Sure[0]:

* Send a proposal to strace. :)

Additionally:

(a) As others have mentioned, it's entirely normal to initially feel like one doesn't understand a new code-base of non-trivial size enough to modify it. One of the skills that participating in a programme like GSoC can provide is learning the tools & strategies that you can use to assist in growing your understanding of new code bases.[1]

(b) From an outside perspective, the fact you're aware of the existence of `strace`, have an interest in contributing to the project & have already poked around in the source code in an attempt to understand it are big positive indicators & differentiate you.

(c) That you're using a tool like `git blame` to assist in your exploration indicates that you've already learned some of the tools/strategies you can use which is also a positive. (As an aside, today I learned about the tool `git-dive` which is intended to be a more powerful form of `git blame`, you might be interested in checking it out: <https://github.com/gitext-rs/git-dive>)

(d) Also, with regard to `strace` specifically, you might gain some insight from this recent video: "strace feels like magic — let's fix that (with Rust)" <https://www.youtube.com/watch?v=engduNoI6DE>. While the video uses Rust for the implementation it also provides a general overview of how `strace` works. (This might be a useful intro to Rust syntax, if you're not currently familiar: <https://fasterthanli.me/articles/a-half-hour-to-learn-rust> (BTW, it is also extremely normal to not understand everything/anything about Rust code on first sight. :) ))

(e) Personally, in the past I've found it extremely easy to talk myself out of submitting proposals for opportunities such as this but my current perspective is: if I have doubts about whether or not I'm experienced/qualified enough to submit a proposal then I'm definitely not qualified to evaluate whether I should submit a proposal--so I definitely should submit a proposal to give those who are qualified the opportunity to evaluate it. (After all, if you don't currently maintain the `strace` project you have no idea what their experience has taught them is important in order for people to successfully contribute.)

Hope some of that is useful.

[0] re: "advices" -- in this context, "advice" is the preferred word[2]. I mention this as written communication is important & based on your previous HN comments this seems to be a pattern of incorrect pluralization rather than a typo. Note: I do not have perfect grammar. :)

[1] Also, I would encourage you to take notes of what you don't understand or issues you ran into when you start exploring a project. You can then document these hurdles/barriers to new contributors for the project so that maintainers are (at least) aware of them. This is one situation where your relative inexperience with a project is in itself valuable because your perspective is one maintainers no longer have. (And such hurdles are important to identify for projects that are actively looking to attract & nurture new contributors.)

[2] <https://en.wiktionary.org/wiki/advices>

mediapipe/docs/solutions/pose.md at master · google/mediapipe · GitHub

Today, we announce MediaPipe diffusion plugins, which enable controllable text-to-image generation to be run on-device. Expanding upon our prior work on GPU inference for on-device large generative models, we introduce new low-cost solutions for controllable text-to-image generation that can be plugged into existing diffusion models and their Low-Rank Adaptation (LoRA) variants.

You can just grab the TFLite version! https://github.com/google/mediapipe/blob/master/docs/solutions/models.md

I believe Google already released transformers under an apache 2 license with a patent grant:

https://github.com/google/mediapipe/blob/master/mediapipe/mo...

I was going to say that I like the MediaPipe Selfie Segmentation model for doing this sort of thing in a web page, but I've just noticed (when getting the GitHub link[1]) that Google have marked the code as legacy[2] ... no idea if the new solution is better/easier to use[3].

For what it's worth, my CodePen using the old model is here: https://codepen.io/kaliedarik/pen/PopBxBM

[1] - https://github.com/google/mediapipe/blob/master/docs/solutio...

[2] - "Attention: Thank you for your interest in MediaPipe Solutions. As of April 4, 2023, this solution was upgraded to a new MediaPipe Solution."

[3] - https://developers.google.com/mediapipe/solutions/vision/ima...

I have used mediapipe very successfully in multiple projects and it's very easy to get running. You can choose from many different vision tasks including hand landmarks ( https://github.com/google/mediapipe/blob/master/docs/solutions/hands.md )

I found MediaPipe's Face Mesh and was impressed with how simple it was to get going, but it just gives you the landmark points and I've not gone any further yet.

We've trained ControlNet on a subset of the LAION-Face dataset using modified output from MediaPipe's face mesh annotator to provide a new level of control when generating images of faces.

Medipaipe library link: https://mediapipe.dev/