IF

AI/ML researcher and CT/Artist base in the UK/EU/JP

[RyanOnTheInside @RyanOnTheInside]

Preview: Stable Audio 3 in real-time with DEMON backend. Coming soon!
Also coming soon, a surface to one-shot vibe code interfaces like the one in this demo video.

[Purz.ai @PurzBeats]

Option paralysis is real. 😱

[ComfyUI @ComfyUI]

Generate split-stack videos like this in a single click.

The basic setup is simple, just upload an image and the system takes care of the rest.

Want more control? Switch to graph mode and fine-tune each step to dial in exactly the look you’re after.

Click the link below for the workflow👇

[IF @ImpactFramesX]

@JesusSavesUs777 Amen

[ComfyUI @ComfyUI]

Ideogram 4.0 is now natively supported on ComfyUI

@ideogram_ai v4.0 is an open-weight 9.3B text-to-image foundation model.
It is exclusively trained on structured JSON caption datasets for precise scene description.

It features flawless text rendering, professional layout control, customizable color palettes, and bounding box adjustment.
Both open-sourced and API node versions are supported in ComfyUI

[ComfyUI @ComfyUI]

The open source community has been delivering on LTX 2.3 LoRAs.

Fine tuning LTX 2.3 unlocks control that you can't get from closed models.

Here are 7 LoRAs that can save your footage.

There are too many incredible LoRAs to cover, so comment below your favorite that needs a showcase!

Links to all the workflows below 👇

[Wang Zhao @WangZhao_0849]

🚀🚀 Introducing Pixal3D (SIGGRAPH’26) — a new pixel-aligned image-to-3D generation paradigm for high-fidelity 3D asset creation.

Today’s Image-to-3D has become pretty good at producing plausible 3D assets. But plausibility is not enough. Fidelity is a hidden bottleneck.
❓A generated model may look “about right,” yet still fail to truly align with the input pixels. Can we make 3D generation as faithful as reconstruction, while still allowing it to complete the unseen?

Pixal3D is our answer.
💡We believe the core bottleneck behind fidelity is 2D–3D correspondence. Most 3D-native generators synthesize shapes in canonical space and inject image cues through cross-attention, forcing the model to implicitly search for which pixels correspond to which 3D regions.

🍀Pixal3D takes a different route. Instead of generating in canonical space, Pixal3D generates directly in pixel-aligned camera space — what you see is what you get. The generated 3D asset is aligned with the input view from the start.
☕️Meanwhile, Pixal3D introduces back-projection-based image condition scheme - explicitly back-projects multi-scale pixel features into 3D voxels, thus resolving the 2D-3D association problem. The input image is no longer just a prompt - it becomes a geometric anchor.

🚩Pixal3D shows that pixel-aligned 3D generation is not only feasible and scalable, but also significantly improves fidelity, pushing 3D-native generation closer to reconstruction-level faithfulness. It also naturally extends to multi-view and scene-level 3D generation.

✅Faithful to the input view. ✅Generative for the unseen.

Closer to reconstruction-level fidelity, with the creativity of 3D generation. Pixal3D also represents an effort towards the unification of 3D generation and reconstruction.

📢Paper, code, and demo are fully released — try it out and let us know your feedback!

🌐Project page: https://t.co/Y1oKzZZrkZ
🤗Huggingface Demo:
https://t.co/4QoDdHMOsk
💻Code:
https://t.co/xwkNNQTMha
📄Paper:
https://t.co/UgiNH00PEY

[Sonia Poulton @SoniaPoulton]

David Attenborough used falsehoods about dead walruses to push a narrative about climate change and then he and Prince William launched the whole lie at the World Economic Forum.

Even when he was corrected by people who actually knew their stuff, he still refused to correct the falsehood.

So, yes, you may be part of the euphoric 'Happy 100th Birthday, Sir David Attenborough - king of nature, blah blah blah' but I think he's a manufactured media character and an establishment apparatchik.

I far and away preferred David Bellamy but he refused to play the game so they ostracised him.

Taken from: The Dark Side of David Attenborough.
Here:
You Tube: https://t.co/L6I5deaWcj
Rumble: https://t.co/HXVcez7ymc
Spotify: https://t.co/c1AqOTIMpS

[jtydhr88 @jtydhr88]

ComfyUI-Mesh2Motion 1.2.0
Custom FBX import is now built into ComfyUI-Mesh2Motion. Load any FBX 3D animation, combine it with professional camera presets, and take precise control over your AI video output.

[青龍聖者 @bdsqlsz]

Z-Image team public D-OPSD:
On-Policy Self-Distillation for Continuously Tuning Step-Distilled Diffusion Models

[Cinema Rhymes 🎬 @CinemaRhymes]

Deleted scenes from In The Mood For Love

[Wildminder @wildmindai]

Omni2Sound.
Yet another unified T2A, V2A, and VT2A.

- silent video > adds sound effects.
- vid + desc > vid with audio.
- vid + audio > syncs precisely while respecting your creative intent

beats AudioX and HY-Foley; supports off-screen synthesis.

https://t.co/tbcOslmZNb

[IF @ImpactFramesX]

@hectorVFX Very good exactly what I need to do.

[Olivio Sarikas @OlivioSarikas]

THIS IS TOO EASY!!!!
Nano Banana Pro + SeeDance 2 FULL GUIDE!

[Brie Wensleydale🧀🐭 @SlipperyGem]

@sunbaolong_2001 has posted an amazing vid about how to use Mesh to Motion and LTX 2.3 together.

Looks pretty great. I'll link his X post below whenever he makes it, he tends to post on Chinese platforms first.
https://t.co/D9PVixbZpb

[有趣的80后程序员 @sunbaolong_2001]

New ComfyUI Extension Alert: Mesh to Motion Explorer. 🧩🚀
This interactive 3D scene editor lets you:
✅ Choose from 124 preset motions (Human, Fox, Bird, etc.)
✅ Align reference images perfectly using Z-Image.
✅ Drive LTX 2.3 via Unicontrol / IC LoRA.
Turning the "uncertainty" of AI into "certainty." Workflow breakdown: 👇
🔗 https://t.co/eUzjLvU8Ya
#ComfyUITutorial #MachineLearning #OpenSourceAI #AITools

[IF @ImpactFramesX]

@sunbaolong_2001 wonderful job once again

[鸟哥 | 蓝鸟会🕊️ @NFTCPS]

Ollama要被干掉了？这个5MB的小东西真的有点东西

有人用Rust写了个叫Shimmy的本地AI推理工具，直接把矛头对准Ollama——

① 单文件5MB，Ollama那边直接哑火
② 启动速度不到100ms，内存只吃50MB
③ OpenAI兼容API，接入成本几乎是零
④ 无需配置、自动分配端口，开箱即用

模型来源也全覆盖了：Hugging Face、Ollama、本地目录，自动识别不用你操心。

体积、速度、内存，三项关键指标全线碾压。想跑本地模型又嫌Ollama太重的，这个值得装一个玩玩。

🔗 https://t.co/nKwBdSYpzO

[VirtuCamera @virtucamera]

After a long wait, VirtuCamera 2 is finally here!!
Now available on Android as well as iOS and now supports Houdini, Cinema 4D, and 3ds Max, joining Blender and Maya
Let’s go! 🚀
#Houdini #Cinema4D #C4D #3dsMax #Blender #b3d #Maya #VFX #VR #mocap #3d #animation #VirtualProduction

[IF @ImpactFramesX]

@ComfyUI @craft_ventures @PaceCap @chemistry Yes, certainly it is. Hard work and making good things, always pays off. But now is time to celebrate.