Eyez_CG

313 posts

Eyez_CG banner
Eyez_CG

Eyez_CG

@EyezCG

all posts are mine only represent me not even my avatar

Montréal, Québec Beigetreten Şubat 2016
167 Folgt118 Follower
Eyez_CG retweetet
机器之心 JIQIZHIXIN
机器之心 JIQIZHIXIN@jiqizhixin·
Xmax X1, the first real-time interactive video model, is here. Powered by autoregressive streaming generation, X1 achieves millisecond ultra-low latency and infinite-length generation, enabling truly natural spatial interaction. Camera is redefined. It’s no longer just a lens, but a magic wand that breaks the barrier between dimensions. Summon virtual beings into your reality and interact with them in real-time. Live implementation video is below, truely amazing work! Definitely one to watch.
Xmax AI@XmaxAIOfficial

The boundary between reality and the virtual world is about to disappear. Imagine characters from Pokemon or Digimon coming to life—not on a screen, but right in front of you. You reach out, and they respond. Introducing X1. The first real-time interactive video model that brings imagination into reality. Built by the Xmax AI team from Tsinghua University. Let’s Play the World through AI. Repost & comment to join the X1 private beta (free).

English
17
116
905
91.5K
Quan Nguyen
Quan Nguyen@qnguyen3·
🚀 Fully offline AI TALKING AVATAR IS HERE! 💻 Stack: • @arcee_ai Virtuoso-Lite (with @ollama ) • Kokoro-80M (TTS) • Whisper.cpp (ASR) 🎭 Streaming to @UnrealEngine Metahuman via Audio2Face 🔥 Running smooth on RTX 4090 Laptop (12GB/16GB) (Excuse the jet engine fans 😅)
English
25
72
584
44.4K
Eyez_CG
Eyez_CG@EyezCG·
@SebAaltonen so will there be a vulkan AI community? As cross platform and optimization oriented as vulkan, at first glance it should be considered by many small businesses or individual devs for migrating with AI. Especially products involve graphics rendering.
English
0
0
0
81
Sebastian Aaltonen
Sebastian Aaltonen@SebAaltonen·
AI companies should hire low-level GPU experts instead of relying solely on high-level frameworks such as Pytorch and TensorFlow. Hand-written CUDA (or even PTX) is still the fastest and most efficient way to drive GPU compute units.
English
41
56
765
41.2K
Eyez_CG retweetet
Min Choi
Min Choi@minchoi·
9. ZooPunk: Use AI to design ships, interact with allies, and experience unique gameplay driven by in-game Stable Diffusion
English
4
8
202
159.6K
Eyez_CG retweetet
Stability AI
Stability AI@StabilityAI·
With a first-of-its-kind architecture, SPAR3D combines precise point cloud sampling with advanced mesh generation to deliver unprecedented control over 3D asset creation. To learn more about the underlying technology, you can read the full research paper on our blog. (3/3)
Stability AI tweet media
English
16
6
41
17.5K
Eyez_CG
Eyez_CG@EyezCG·
@__JAAF__ did you use the model on comfyui or?
English
0
0
0
10
JAAF
JAAF@__JAAF__·
A quick test of Stable Point aware 3D on Hugging Face 🤗 From image to 3d mesh in seconds. Impressed with the speed in particular
English
1
0
3
302
Eyez_CG retweetet
BensenHsu
BensenHsu@BensenHsu·
The paper introduces Genie 2, a large-scale foundation world model that can generate an endless variety of action-controllable, playable 3D environments. This is intended to enable future AI agents to be trained and evaluated in a limitless curriculum of novel worlds. Genie 2 demonstrates various emergent capabilities, such as object interactions, complex character animation, physics simulation, and the ability to model the behavior of other agents. It can generate consistent worlds for up to a minute and supports diverse perspectives like first-person, isometric, and third-person views. The authors suggest that Genie 2 could enable future agents to be trained and evaluated in a limitless curriculum of novel worlds, overcoming the traditional bottleneck of available training environments. It also enables rapid prototyping of diverse interactive experiences, which can accelerate the creative process for environment design and research. full research: openread.academy/en/paper/readi…
BensenHsu tweet media
English
1
5
37
13.1K
Daniel Skaale
Daniel Skaale@DSkaale·
@EyezCG @NianticLabs Oh I did scan a part from a old submarine with the Scaninvers app but you are absolutely right I should not have used this tag 🙂
English
1
0
0
63
Eyez_CG
Eyez_CG@EyezCG·
@ko_yuki_lo 可愛い!予定させていただきませんか🫶🏼
日本語
1
0
0
18
YukiYa 🇨🇭→🇯🇵→🇨🇦
20個のUVノードのキーホルダーが届いております!これから液体挿入、パケージできるような材料検討と購入します。出来次第、前もってショップ開店予定日をお伝えします。
YukiYa 🇨🇭→🇯🇵→🇨🇦 tweet media
日本語
3
0
19
580
Harry Nesbitt 🐑
Harry Nesbitt 🐑@harrynesbitt·
I've recently been working on my own realtime GI system for Unity, inspired by the work of @CasualEffects. It uses software raytracing to gather light into a grid of probes, which any surface can sample. Zero baking, no light leaks, infinite bounces! #madewithunity #gamedev
English
36
193
1.3K
0
Eyez_CG retweetet
Jim Fan
Jim Fan@DrJimFan·
Exciting updates on Project GR00T! We discover a systematic way to scale up robot data, tackling the most painful pain point in robotics. The idea is simple: human collects demonstration on a real robot, and we multiply that data 1000x or more in simulation. Let’s break it down: 1. We use Apple Vision Pro (yes!!) to give the human operator first person control of the humanoid. Vision Pro parses human hand pose and retargets the motion to the robot hand, all in real time. From the human’s point of view, they are immersed in another body like the Avatar. Teleoperation is slow and time-consuming, but we can afford to collect a small amount of data. 2. We use RoboCasa, a generative simulation framework, to multiply the demonstration data by varying the visual appearance and layout of the environment. In Jensen’s keynote video below, the humanoid is now placing the cup in hundreds of kitchens with a huge diversity of textures, furniture, and object placement. We only have 1 physical kitchen at the GEAR Lab in NVIDIA HQ, but we can conjure up infinite ones in simulation. 3. Finally, we apply MimicGen, a technique to multiply the above data even more by varying the *motion* of the robot. MimicGen generates vast number of new action trajectories based on the original human data, and filters out failed ones (e.g. those that drop the cup) to form a much larger dataset. To sum up, given 1 human trajectory with Vision Pro -> RoboCasa produces N (varying visuals) -> MimicGen further augments to NxM (varying motions). This is the way to trade compute for expensive human data by GPU-accelerated simulation. A while ago, I mentioned that teleoperation is fundamentally not scalable, because we are always limited by 24 hrs/robot/day in the world of atoms. Our new GR00T synthetic data pipeline breaks this barrier in the world of bits. Scaling has been so much fun for LLMs, and it's finally our turn to have fun in robotics! We are building tools to enable everyone in the ecosystem to scale up with us. Links in thread:
English
73
342
1.8K
364.1K
Eyez_CG
Eyez_CG@EyezCG·
@AnkaHeChen amazing work! frame time means time to render per frame? (coming from real time graphics field)
English
0
0
0
45
Eyez_CG
Eyez_CG@EyezCG·
@ko_yuki_lo めっちゃ欲しいです!im also at montreal haha, are you selling them?
日本語
1
0
1
17
YukiYa 🇨🇭→🇯🇵→🇨🇦
届いた!! 今夜液体を入れてみるよ。 サイズは一回り小さくていいかもね😅
YukiYa 🇨🇭→🇯🇵→🇨🇦 tweet media
日本語
18
86
912
48.2K
Eyez_CG retweetet
brett goldstein
brett goldstein@thatguybg·
Sora's video quality seems impossible so I dug into how it works under the hood it uses both diffusion (starting with noise, refining towards a desired video) and transformer architectures (handling sequential video frames) read on 🧵
GIF
English
95
772
5.8K
1.7M
Eyez_CG
Eyez_CG@EyezCG·
@Michael_Moroz_ hahah thats cool! id love to! i added u on both vrc and discord
English
1
0
1
46
Mykhailo Moroz
Mykhailo Moroz@Michael_Moroz_·
@EyezCG unfortunately the only way to do so is to visit me in vrchat
English
1
0
1
57
Mykhailo Moroz
Mykhailo Moroz@Michael_Moroz_·
Implemented anisotropic kernels in my VRChat fluid sim (cs.nyu.edu/exact/doc/anis…) Now to figure out how to get high quality normals out of this... (no screen space, thanks) Also still need to make tighter billboard quads for the ellipsoids, rn pretty expensive to render
English
11
74
535
53.7K