search

HumanAIGC / AnimateAnyone

Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation
Apache License 2.0
14.15k stars 942 forks source link

Can the developer(s) perhaps give us weekly updates on the status of the project for public transparency on what's happening? #48

Open Collin-Budrick opened 7 months ago

Collin-Budrick commented 7 months ago

I recommend creating a Discord server for the project and setting up an announcement channel like a lot of AI projects do.

ReEnMikki commented 7 months ago

The Issues section of this repo is already such a cesspool, I can't imagine the hellhole that is called the Discord server for this project 💀💀💀

Also looks like this repo has reached its plateau. The rate of it gaining stars with zero code has practically became minimal now. 10k+ ⭐ with no code is always funny.

Collin-Budrick commented 7 months ago

The Issues section of this repo is already such a cesspool, I can't imagine the hellhole that is called the Discord server for this project 💀💀💀

Yeah, well, the authors can't give a date when they will release the code, and they don't put out updates routinely. I'm beginning to believe the rumors that this was at least partly fake. It shouldn't take this long to polish up the code for ease of use.

There's a number of people on this repo who are already putting the paper into code by themselves. Kinda embarrassing/suspicious at this point that they haven't even released the model at this point.

cepera-ang commented 7 months ago

It is not up to authors only to decide to publish corporate intellectual property. Especially after public stir that happened here. You may be pretty sure that every corporate function now feels very averse to ever allowing this code to be published. First, it is clearly seem more valuable now, given the reaction of disbelief. Second, it is now risky that research code will be treated with expectations of a polished product. Third, all the SJW activity around model likeness use.

So, quite foreseeable if this code release will switch from "interns gonna publish when they feel comfortable" to "Alibaba will publish it when and if legal and marketing and big bosses feel comfortable and if deemed necessary".

On Tue, Dec 19, 2023, 14:03 Collin Budrick @.***> wrote:

The Issues section of this repo is already such a cesspool, I can't imagine the hellhole that is called the Discord server for this project 💀💀💀

Yeah, well, the authors can't give a date when they will release the code, and they don't put out updates routinely. I'm beginning to believe the rumors that this was at least partly fake. It shouldn't take this long to polish up the code for ease of use.

There's a number of people on this repo who are already putting the paper into code by themselves. Kinda embarrassing/suspicious at this point that they haven't even released the model at this point.

— Reply to this email directly, view it on GitHub https://github.com/HumanAIGC/AnimateAnyone/issues/48#issuecomment-1862226428, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACRP6OYZLJ4C2SFEH242CDLYKE34BAVCNFSM6AAAAABA2BBFXSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNRSGIZDMNBSHA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

Collin-Budrick commented 7 months ago

@cepera-ang they already committed in their statement where they were with the process... at the very least, they should be transparent about what they're up to through weekly updates. Even if it's ethical concerns or corporate red tape. You can't publish a demo, garnish attention, and say, "well, we will publish it eventually, just hang tight and don't ask us when... also cya next year."

please be certain that the intention to provide access to both the demo and our source code is firm.

We are past the red tape.

chrisbward commented 6 months ago

Stop being "impatient and shallow-minded"? You turned quickly. Call it how I see it.

Collin-Budrick commented 6 months ago

Stop being "impatient and shallow-minded"? You turned quickly. Call it how I see it.

Context is key... im not accusing them of fake work, I'm merely asking for updates on the progress.

lucasjinreal commented 6 months ago

Please do not hope a none existed project opensource, this is just a PR. Don't be too serious.

Collin-Budrick commented 6 months ago

Please do not hope a none existed project opensource, this is just a PR. Don't be too serious.

Then why the f would they post that update on the read me?

lucasjinreal commented 6 months ago

It's Chinese style. Lots of README like opensource out there for Chinese companies.

continue-revolution commented 6 months ago

Some previous projects from Alibaba have a 2-month gap between release and open-source because they have a long internal review process. Actually a lot of Chinese companies / research labs have. So let's wait patiently.

Collin-Budrick commented 4 months ago

@continue-revolution after looking at this authors other projects, there doesn't seem to be any code... just baseless marketing nonsense.

continue-revolution commented 4 months ago

Yes this group seems to be releasing nothing. Different Alibaba lab have different style. I guess this one is like, well let’s just create a blank repo and see if it can be hot, if it is then let’s make it a product, otherwise let’s open-source. DragonDiffusion (not from Alibaba) takes 7 months to open-source, so who knows. Well now this project has already become a product in their own TongYi app and no one knows if they gonna open-source

ShawnFumo commented 3 months ago

Speaking of diff Alibaba labs, a different paper for animating people called CHAMP was just released, and already has some code and weights available: https://github.com/fudan-generative-vision/champ