3DeSocial.deso (@3DeSocial) header
3DeSocial.deso (@3DeSocial) profile picture

Mixing AI with spatial and metaverse for the decentralized social media future. The latest spatial, metaverse and gaming news in our daily Spatial Tech Update.

Joined August 2021
Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)Photo by 3DeSocial.deso (@3DeSocial)
3DeSocial.deso retweeted
Today, we introduce HunyuanImage 3.0-Instruct, a native multimodal model focusing on image-editing by integrating visual understanding with precise image synthesis! πŸš€ It understands input images and reasons before generating images. Built on an 80B-parameter MoE architecture (13B activated), it natively unifies deep multimodal comprehension and high-fidelity generation. 🧠 A "Thinking" Model with Native CoT & MixGRPO: The model doesn’t just execute commands, it processes them through a Native Chain-of-Thought (CoT) schema. Enhanced by our self-developed MixGRPO algorithm, it reasons through complex instructions to achieve flawless intent alignment and human-preference consistency. 🎨 Precise Editing & Multi-Image Fusion: The model enables accurate image editing by adding, removing, or modifying elements while keeping non-target areas perfectly intact. It also excels at seamless multi-image fusion, synthesizing complex scenes by extracting and blending elements from multiple sources into a unified, consistent output. πŸ† SOTA Performance: HunyuanImage 3.0-Instruct sets a new benchmark in visual quality and alignment, delivering performance that matches leading proprietary models. We aim to enable the community to explore new ideas with a state-of-the-art foundation model, fostering a dynamic and vibrant image generation ecosystem. πŸ› οΈπŸŽ¨ πŸ’»Try it at (PC only): hunyuan.tencent.com/chat/Hun…
44
140
37
1,249
127,172
3DeSocial.deso retweeted
Image to 3D, now in ANY pose! πŸ€Έβ€β™‚οΈ Simply upload your image + a pose reference photo. Meshy generates the model to match it perfectly. No complex rigging needed. Create 3D characters your way.
40
113
31
1,002
103,903
3DeSocial.deso retweeted
Anime.js website is nominated for Site of the Year on @awwwards! Lots of great contenders this year! If you'd like to support the project, your vote would mean a lot! (Link below β–Ό)
42
95
5
1,514
85,332
3DeSocial.deso retweeted
You can turn any Google Street View into Video. Anywhere on Earth, using nothing but your own PC. I just generated a walk through Bryant Park, New York.
8GB is All You Need --- Generate AI Videos with LTX-2 AI Model on Your PC! WanGP is a dead simple Gradio app for generating AI videos, and you can run the new LTX-2 model. And now: - Just 8GB VRAM required! - High-res videos with lower VRAM! Run it with 1-click on Pinokio
15
74
4
931
173,733
3DeSocial.deso retweeted
Mainstream media tends to think VR headset growth is slowing down, and the metaverse hype has cooled off. Yet VRChat hit a new all time high on New Year’s, with 149,000 concurrent users. Yes, the core VR user base is still steadily expanding. One special thing about VRChat is that the ceiling of its user experience is extremely high. By contrast, over the past few years, some similar products have often fallen into two traps. The first is adapting to mobile too early. Mobile is a mature platform, so in the short term you can definitely gain a large number of users. But to fit mobile, you inevitably make major compromises in performance and interaction, which caps how far the VR experience can go. The second is lowering the creation barrier. A lower barrier obviously means more creators and more content. But the lower the barrier, the more you usually sacrifice freedom and the experience ceiling, and the quality of what gets made tends not to be as high. Both strategies can look great in the short term if you only look at the numbers, but in the medium to long run they are not helpful for shaping a truly compelling VR experience. VRChat focused on serving its most hardcore PC VR users first. It launched a VR build back in 2014 for the DK1 era, and didn’t release a mobile app beta until 2023. Even today, many worlds on VRChat are still PC VR only. On the creator side, VRChat supports Unity, and the learning curve is relatively high. Of course, many people also criticize the team for being lazy, moving too slowly, and having loose management. But I think that’s exactly what allowed the most core users to create worlds and avatars with near total freedom and almost no ceiling, and to visit those VR worlds on the best possible setups. That may be the reason VRChat has remained so resilient all these years.
Happy New Year, MT! Our new CCU ATH is 148,886, set during US CT (UTC-6)! Fully stable throughout. Big kudos to my #VRChat teammates & our service providers. Stick around for PT NYE!
9
9
125
13,947
3DeSocial.deso retweeted
AI aided 3D world building is getting ridiculous. This is a large, multi-room gaussian splat (~6million splats). Has streaming, LoD, character support, spatial audio, physics, collisions, VR support, and streaming video. Audio on ... Great place to watch a movie :)
41
59
6
590
47,742
3DeSocial.deso retweeted
You can now generate and animate 3D characters in < 5mins with AI! –Nano Banana Pro "Generate a T-pose of <anything>" –Hunyuan3D 3.1: image to 3D model with textures (.obj / .fpx) –Mixamo: 3D humanoid model rigging –Claude: use three.js to render Try it with a pic you upload!
145
340
32
3,610
365,161
3DeSocial.deso retweeted
VR stress test (recording was done on mac, but works well on Q3) - Universe of ~500m splats - Collisions at scale - Physics - Character insertion You're going to be able to build some crazy shit for VR soon ...
11
5
1
113
17,037
3DeSocial.deso retweeted
Today I'm releasing PortalVR 2.0, which lets you play the entire SteamVR ecosystem without wearing a headset. You plug in your Quest, put it on your desk, and play on your monitor. It supports 3D glasses and even 3D monitors! 😎 Get it now: portalvr.io Thread 🧡
211
367
233
3,735
2,320,224
Congratulations to @MoonHoodStudios on winning vr game of the year with The Midnight Walk. I’ve noticed a spike in views on my review since this happened tooπŸ‘
14
7
1
105
2,788
3DeSocial.deso retweeted
oh my.. hows this even possible Meta just dropped SAM 3D, you can auto select any object in still image and.. turn them into high quality 3D model the model is free to download and test in playground link in comments
196
1,015
145
11,848
796,698
3DeSocial.deso retweeted
Occasionally I try an AI tool so incredible that I simply can’t stop playing with it. The new @theworldlabs model definitely qualifies here. It’s almost impossible to describe how cool it is to walk through a 3D world that you generated from a single image.
Generate persistent 3D worlds from a single image, bigger and better than ever! We’re excited to share our latest results and invite you to try out our world generation model in a limited beta preview.
65
189
22
2,151
273,173
3DeSocial.deso retweeted
There's only one chain that can achieve this feat. Eventually information will be stored on-chain, which will allow public auditing & analysis by anyone in the world to decipher truth. This is way bigger than money.
People understand - if the official news, images and video were placed onchain, then you will read the truth. The blockchain networks will rebuild truth. This is way bigger than money.
10
7
48
6,815
3DeSocial.deso retweeted
every day we stray further from god
279
2,401
98
42,258
920,316
Grok Code Fast 1 is now available for free in Windsurf for a limited time. Let us know what you think!
74
77
61
917
178,189
3DeSocial.deso retweeted
This entire world was built in realtime and collaboratively by people who simply prompted what they wanted in the chat and a few seconds later it appeared like magic πŸͺ„ It runs on mobile, tablet, desktop and VR by literally clicking a link and loads instantly. Soon, you'll be able to voice prompt AI in VR and grab and move things around 😈 The holodeck isn't coming, it's already here ⚑️
16
35
9
169
25,001
folks this stuff in the VR browser feels amazing. watching @GertJanAkerboom prompting a thing bear... 🐻@hyperfy_io #webXR @threejs
6
14
4
65
2,673
SkyworkAI released Matrix-3D which can generate 3D worlds from a single image or text prompt:
6
17
1
181
13,990
3DeSocial.deso retweeted
πŸ΄β€β˜ οΈ Welcome to the IMAX era of AI 3D. Introducing TRIPO 3.0. It's time to say goodbye to blurriness and take full control of your details. #TripoAI #TechForCreators #Tripostudio #3DModeling
17
73
21
404
130,501
3DeSocial.deso retweeted
Good luck trying this on DeSo where anyone in the world can run a node, and build an app with protocol-level encrypted messaging, that’s all fully open-sourced.
7
11
1
40
5,676