Tianyu (Steve) Wang

Research Scientist @AdobeResearch | Ph.D. @ CUHK | Prev. Research Intern @AdobeResearch | Photographer | Opinions are my own

[Tianyu (Steve) Wang @VisionSteve]

Check out our project that answers how to train any-step T2I model from scratch.

We release the code for everyone to explore this area.

Looking forward to seeing more on this!!

#CVPR #CVPR2026

[Xin Yu (Andy) @andy_yx27]

😀😀We’re excited to release the training code for Self-E (accepted to CVPR 2026): https://t.co/Fy1T6EnEpO

Self-E is a training-from-scratch, self-contained framework for any-step text-to-image generation, without teacher distillation.

[Tianwei Yin @TianweiY]

1/ Our new @reve image model is now #2 on the @arena text-to-image leaderboard — behind only GPT Image 2, ahead of Nano Banana Pro, Microsoft, xAI and everyone else.

And it's a 125 point jump over Reve 1.5 from just 3 months ago.

The research story behind it 🧵👇

[Tianyu (Steve) Wang @VisionSteve]

@TianweiY @reve @arena So coool!

[Yukang Chen @yukangchen_]

🚀 Excited to release LongLive 2.0!
🎬 An end-to-end infrastructure for long video generation, with FP4 and parallelism at the core of both training and inference.

⚡45.7 FPS generation speed on 5B model⚡

✨ LongLive 2.0 supports real-video training, few-step distillation, multi-shot training/inference, sequence-parallel acceleration, NVFP4 KV cache, and async VAE decoding deployment.

🧩 To our knowledge, this is the first open-source 4-bit long video generation infra that covers both training and inference.

🙌 Welcome to check it out, try it, and share feedback!

🔗 Code: https://t.co/QXF2lfnNzL
📰 Paper: https://t.co/gKtarHj17c
🎥 Demo: https://t.co/RLF1wfOXVZ

#LongVideoGeneration #VideoGeneration #Realtime #AIInfra #EfficientAI #FP4 #Parallel #NVIDIA

[Tianyu (Steve) Wang @VisionSteve]

@jin_linyi @holynski_ So cool!!!

[Tianyu (Steve) Wang @VisionSteve]

@xiaolonw @Lianhuiq @Meta Congrats!

[Tianyu (Steve) Wang @VisionSteve]

RT @BoyuanChen0: If there is one secret recipe about GPT Image 2, it must be codex. In the last 6 months, codex has been zero-shoting every…

[Tianyu (Steve) Wang @VisionSteve]

If you’re attending, please stop by our oral session for EditVerse:

I won’t be there in person, but I’d be very happy if you could check out the work, chat with the team, and join the discussion.

📍 Room 201 A/B
🕚 Apr 25, 11:18–11:28 AM

[Xuan Ju @juxuan_27]

Excited to share our paper EditedVerse is accepted as oral to ICLR 2026! Many thanks to our amazing coauthors!!

Paper Link: https://t.co/9B0BoCE2zH
Project Page: https://t.co/FPQ9hTG0qc

[Xuan Ju @juxuan_27]

Excited to share our paper EditedVerse is accepted as oral to ICLR 2026! Many thanks to our amazing coauthors!!

Paper Link: https://t.co/9B0BoCE2zH
Project Page: https://t.co/FPQ9hTG0qc

[GZ Zhou @GengzeZhou]

We present LightMover — controllable light movement from a single image.

Move lights. Change color. Adjust intensity.
All with physically consistent shadows & reflections.
🌐 Project: https://t.co/NdRLXM8PsZ
📄 Paper: https://t.co/EY40m9cHvb

#CVPR2026

[Xin Yu (Andy) @andy_yx27]

😀😀We’re excited to release the training code for Self-E (accepted to CVPR 2026): https://t.co/Fy1T6EnEpO

Self-E is a training-from-scratch, self-contained framework for any-step text-to-image generation, without teacher distillation.

[Tianyu (Steve) Wang @VisionSteve]

@natanielruizg @JaredSleeper It would be good to say “I will finish your job, please go to sleep” and then finish the job 🤣

[Tianyu (Steve) Wang @VisionSteve]

@natanielruizg @CSProfKGD Great work!!

[Xun Huang @xxunhuang]

This is the first video world model to support multi agent interactions, a truly groundbreaking milestone. Awesome work by @sainingxie and the team! Excited to see Self Forcing powering multiplayer world models as well.

[Xin Yu (Andy) @andy_yx27]

Excited to share our new work Self-E: A New Training Paradigm for Text-to-Image!

One model, any compute: Unlock any-step text-to-image generation. Fully trained from scratch, no teacher distillation needed.
https://t.co/cyfnmMKHP7

The secret? Let the model evaluate itself. 👇

[Yao-Chih Lee @YaoChihLee]

Excited to share our new work: Generative Video Motion Editing with 3D Point Tracks.

We propose a framework that uses 3D point tracks to precisely edit both camera and object motion in a video, unlocking a wide range of new editing applications.

[Tianyu (Steve) Wang @VisionSteve]

The last version of FSD V13 is good! Confident and smooth. But the v14.1.4 is bad, not confident and not smooth. I put my hand back to the wheel again.😅 Cancel subscription again and wait for a stable version.

[Andrej Karpathy @karpathy]

I took delivery of a beautiful new shiny HW4 Tesla Model X today, so I immediately took it out for an FSD test drive, a bit like I used to do almost daily for 5 years. Basically... I'm amazed - it drives really, really well, smooth, confident, noticeably better than what I'm used to on HW3 (my previous car) and eons ahead of the version I remember driving up highway 280 on my first day at Tesla ~9 years ago, where I had to intervene every time the road mildly curved or sloped. (note this is v13, my car hasn't been offered the latest v14 yet)

On the highway, I felt like a passenger in some super high tech Maglev train pod - the car is locked in the center of the lane while I'm looking out from Model X's higher vantage point and its panoramic front window, listening to the (incredible) sound system, or chatting with Grok. On city streets, the car casually handled a number of tricky scenarios that I remember losing sleep over just a few years ago. It negotiated incoming cars in tight lanes, it gracefully went around construction and temporarily in-lane stationary cars, it correctly timed tricky left turns with incoming traffic from both sides, it gracefully gave way to the car that went out of order in the 4-way stop sign, it found a way to squeeze into a bumper to bumper traffic to make its turn, it overtook the bus that was loading passengers but still stopped for the stop sign that was blocked by the bus, and at the end of the route it circled around a parking lot, found a spot and... parked. Basically a flawless drive.

For context, I'm used to going out for a brief test drive around the neighborhood to return with 20 clips of things that could be improved. It's new for me to do just that and exactly like I used to, but come back with nothing. Perfect drive, no notes. I expect there's still more work for the team in the long march of 9s, but it's just so cool to see that we're beyond finding issues on any individual ~1 hour drive around the neighborhood, you actually have to go to the fleet and mine them. Back then, I processed the incredible promise of vehicle autonomy at scale (in the fully scaleable, vision only, end-to-end Tesla way) only intellectually, but now it is possible to feel it intuitively too if you just go out for a drive. Wait, of course surround video stream at 60Hz processed by a fully dedicated "driving brain" neural net will work, and it will be so much better and safer than a human driver. Did anyone else think otherwise?

I also watched @aelluswamy 's new ICCV25 talk last week (https://t.co/RdaM23kvez) that hints at some of the recent under the hood technical components driving this progress. Sensor streams (videos, maps, kinematics, audio, ...) over long contexts (e.g. ~30 seconds) go into a big neural net, steering/acceleration comes out, optionally with visualization auxiliary data. This is the dream of the complete Software 1.0 -> Software 2.0 re-write that scales fully with data streaming from millions of cars in the fleet and the compute capacity of your chip, not some engineer's clever new DoubleParkedCarHandler C++ abstraction with undefined test-time characteristics of memory and runtime. There's a lot more hints in the video on where things are going with the emerging "robotics+AI at scale stack". World reconstructors, world simulators "dreaming" dynamics, RL, all of these components general, foundational, neural net based, how the car is really just one kind of robot... are people getting this yet?

Huge congrats to the team - you're building magic objects of the future, you rock! And I love my car <3.

[Umesh @umesh_ai]

🚨 UNLIMITED EVERYTHING for 30 days!

From Oct 28–Dec 1, every Adobe Creative Cloud & Firefly subscriber unlocks:

✨ Unlimited access to Firefly & partner AI models (ChatGPT, Ideogram 3.0, nano banana, Flux, Firefly 5 and more).

✨ Even video generation is unlimited in Relax mode.

No credits. No limits.

Let your imagination run wild → I’ve linked some of my favorite prompts and tips to get started.

Sponsored by @Adobe as an Adobe Firefly Ambassador.

#AdobeFireflyAmbassadors #Ad #FireflyPromo

[Xun Huang @xxunhuang]

We present MotionStream — real-time, long-duration video generation that you can interactively control just by dragging your mouse.

All videos here are raw, real-time screen captures without any post-processing. Model runs on a single H100 at 29 FPS and 0.4s latency.

[Tianyu (Steve) Wang @VisionSteve]

@redlilyrose19 thanks!!

[Tianyu (Steve) Wang @VisionSteve]

Very proud that two of the projects I worked on were featured in this year’s Adobe MAX Sneaks!
I co-led Project Frame Forward, which builds upon our previous GenProp — with major improvements in stability and image editing &amp; video alignment. #AdobeMax #ProjectFrameForward

[Jerrod Lew @jerrod_lew]

AI video editing on show here at Adobe Max!

#ProjectFrameForward lets you alter the start frame but keep the same motion of the original video.

Here’s a quick demo: