Joined February 2013
50 Photos and videos
Pinned Tweet
LongAnimateDiff: Our new AnimateDiff model that supports up to 64 frames! We are excited to announce the launch of our advanced motion module, designed for long video generation. This cutting-edge module, developed based on the #AnimateDiff model (v2) for Stable Diffusion, now enables video generation of up to 64 frames, surpassing the industry's current maximum of 25 frames set by @stability.ai's #SVD. Moreover, our model offers enhanced flexibility, allowing for a variety of video lengths ranging from 16 to 64 frames. For creators and developers, the model is now available in #ComfyUI. code: github.com/Lightricks/LongAnโ€ฆ @HuggingFace: huggingface.co/Lightricks/Loโ€ฆ @s_weissbuch Daniel Shalem @yoavhacohen @ofirbb @Lightricks
11
34
147
67,065
love to see our paper working so nice! arxiv.org/abs/2601.22143
Tested @LTX's new LipDub capability on Tarantino's strudel scene - 11 languages, one model, dramatic close-ups preserved. This is what streaming localization looks like with the new generation of AI video models. ๐Ÿ‡ซ๐Ÿ‡ท๐Ÿ‡ท๐Ÿ‡บ ๐Ÿ‡ฉ๐Ÿ‡ช ๐Ÿ‡ช๐Ÿ‡ธ ๐Ÿ‡ต๐Ÿ‡น ๐Ÿ‡ฏ๐Ÿ‡ต ๐Ÿ‡ฎ๐Ÿ‡น ๐Ÿ‡จ๐Ÿ‡ณ ๐Ÿ‡ฌ๐Ÿ‡ง ๐Ÿ‡ฎ๐Ÿ‡ฑ ๐Ÿ‡ฐ๐Ÿ‡ท ๐Ÿ‡ธ๐Ÿ‡ฆ
1
18
2,346
I love seeing our work @MedOumoumad out there all around the world!
็”Ÿๆˆใ—ใŸๅ‹•็”ปใ‚’HDRๅŒ–ใ™ใ‚‹LTX-2.3ใฎHDR LoRAใฎใƒ†ใ‚นใƒˆ EXRใงๆ›ธใๅ‡บใ›ใ‚‹ใ‹ใ‚‰ๆ’ฎๅฝฑ็ด ๆใ‚„CGใจ้ฆดๆŸ“ใพใ›ใŸใ‚Šใ‚ฐใƒฌใƒผใƒ‡ใ‚ฃใƒณใ‚ฐใŒๅฏ่ƒฝใซใชใ‚‹ใ‚„ใค ใƒญใƒผใ‚ซใƒซใงใ‚‚ๅ‹•ใใ—็™ฝ้ฃ›ใณใ‚‚ๅพฉๅ…ƒใ•ใ‚Œใฆใฆๅ“่ณชๆ‚ชใใชใ„ใฎใงใ“ใ‚Œใฏๅธธ็”จ
1
6
558
It was a big pleasure to work on this together with @antonch44379293 and Gal Zeevi!
6
392
1/2 Excited to share my work with @antonch44379293 & Gal Zeevi: Just Dub It! ๐ŸŽฌ๐Ÿ”Š We solve dubbing in one pass via joint audio/video on @ltx_model, handling extreme profiles & occlusions with ease. Accepted to SIGGRAPH 2026! ๐ŸŽ“ ๐Ÿงต๐Ÿ‘‡
11
8
44
7,043
This was a fantastic collaboration with @DanielCohenOr1 and @OPatashnik Weโ€™re proud to open-source everything today for the community! ๐ŸŒช๏ธ ๐ŸŒ Project: justdubit.github.io/ ๐Ÿ“„ Paper: arxiv.org/abs/2601.22143 ๐Ÿ’ป Code: github.com/Lightricks/LTX-2/โ€ฆ ๐Ÿค— Model: huggingface.co/Lightricks/LTโ€ฆ

1
449
1/5 Excited to share our latest: Bringing HDR to Generative Video! ๐Ÿš€ The vision started with @AmirGa, who realized LTX should support HDR. Through a joint effort, @MedOumoumad and I brought it to life. ๐Ÿ”— arxiv.org/abs/2604.11788
3
9
34
2,461
4/5 Thatโ€™s the beauty of the generative domain: formulate the problem in the right space, and you get amazing results with very little data. Proud to go from zero expertise to a full Beta in just 8 weeks! ๐ŸŽฌโœจ
1
1
124
5/5 An incredible academic and professional collaboration. Huge thanks to Prof. @DanielCohenOr1 and Prof. @OPatashnik for their guidance on the paper, and the whole team for pushing this through! ๐Ÿค #GenerativeAI #HDR #AIResearch #LTX #MachineLearning
2
127
Creators here! Me & @MedOumoumad used @LTX to fix "camera corruptions" and unlock true HDR. ๐Ÿš€ Check out also the project page : hdr-lumivid.github.io/

What this means is AI models that generate video are about to give us much better quality. But this line caught my eye: โ€œWe teach the model to recover those details by training it through deliberate corruption.โ€ I love reading papers like this. Shows how the future gets better. And has creative thinking and engineering hidden inside.
2
235
Naomi Ken Korem retweeted
What this means is AI models that generate video are about to give us much better quality. But this line caught my eye: โ€œWe teach the model to recover those details by training it through deliberate corruption.โ€ I love reading papers like this. Shows how the future gets better. And has creative thinking and engineering hidden inside.
8
6
72
7,699
Naomi Ken Korem retweeted
๐ŸŽ™๏ธ Introducing ID-LoRA: the first open-source model to jointly generate a video with a person's appearance and voice in a single pass from just a reference image short audio clip. No more cascaded pipelines where the audio can't follow your prompt. youtu.be/6bWcMh18K6g?si=_Dgjโ€ฆ
3
11
23
2,178
ืฆืขื“ื ื• ืืชืžื•ืœ ื‘ืฉื‘ื™ืœ ื‘ื™ื˜ื—ื•ืŸ ื”ืžื“ื™ื ื”. ืฆืขื“ื ื• ื›ื“ื™ ืœื”ื–ื›ื™ืจ ืœื›ื•ืœื. ื‘ื™ื˜ื—ื•ืŸ ื”ืžื“ื™ื ื” ื”ื•ื ืžืขืœ ื”ื›ืœ. ืื™ืŸ ืžืงื•ื ืœืžืชื•ื•ื” ืฉื™ื•ื•ืงื™. ืฉื ืชื™ื™ื ืœืชื•ืš ืœื—ื™ืžื”, ืฆืจื™ืš ื—ื•ืง ืฉื™ื‘ื™ื ืœื’ื™ื•ืก ื ืจื—ื‘ ืžื”ื™ืจ ื•ืืคืงื˜ื™ื‘ื™. ื”ืžื’ื–ืจ ื”ื—ืจื“ื™ ืฆืจื™ืš ืœื—ื–ื•ืจ ืœื”ื™ื•ืช ืฉื•ืชืฃ ื‘ื”ื’ื ื” ืขืœ ื”ืžื“ื™ื ื”.
ื’ื™ื•ืก ื”ื•ื’ืŸ ื•ืืžื™ืชื™ ืœื›ื•ืœื ื”ื•ื ืฆื• ื”ืฉืขื” ๐Ÿ‡ฎ๐Ÿ‡ฑ ืฆืขื“ื ื• ื”ืขืจื‘ ื™ื—ื“, ืžืฉืคื—ื•ืช ืžื™ืœื•ืื™ืžื ื™ืงื™ื ื•ืžืฉืคื—ื•ืช ืฉื›ื•ืœื•ืช, ื“ืชื™ื™ื ื•ื—ื™ืœื•ื ื™ื ื›ืื—ื“, ื›ื“ื™ ืœื“ืจื•ืฉ ื—ืœื•ืงื” ืฆื•ื“ืงืช ื•ื ื›ื•ื ื” ื™ื•ืชืจ ืฉืœ ื”ื ื˜ืœ ืœืฉืžื™ืจื” ืขืœ ื‘ื˜ื—ื•ื ื” ืฉืœ ืžื“ื™ื ืช ื™ืฉืจืืœ. ื”ืžืฆื‘ ื‘ื• ื›ืœ ื ื˜ืœ ื”ืœื—ื™ืžื” ืžื•ื˜ืœ ืขืœ ืงื‘ื•ืฆื” ืกืคืฆื™ืคื™ืช ืฉืœ ืื•ืชื ื”ืื ืฉื™ื ืฉื ืงืจืื™ื ืฉื•ื‘ ื•ืฉื•ื‘ ื•ืžืฉืœืžื™ื ืžื—ื™ืจ ืื“ื™ืจ, ื‘ื–ืžืŸ ืฉืื—ืจื™ื ื–ื•ื›ื™ื ืœืดืคื˜ื•ืจืด - ืœื ื™ื›ื•ืœ ืœื”ื™ืžืฉืš. ื ืžืฉื™ืš ืœื”ื™ืื‘ืง ื™ื—ื“ ื›ืœ ื”ื“ืจืš ืœืžืขืŸ ื’ื™ื•ืก ืืžื™ืชื™ ื•ืฆื•ื“ืง.
2
97
Naomi Ken Korem retweeted
29 Dec 2025
ื—ื ื ื™ื” ื‘ื™ืŸ ืฉืžืขื•ืŸ @hananiabsh ืžืคืจืง ื˜ื™ืขื•ืŸ ืกืจืง ืขืœ ืฆืจื›ื™ ื›ื•ื— ื”ืื“ื ืฉืœ ืฆื”"ืœ: "ืœืžื” ืœื ืžื’ื™ื™ืกื™ื ืื ืฉื™ ืžื™ืœื•ืื™ื ืฉืงื™ื‘ืœื• ืคื˜ื•ืจ?" #ืžื”ืฆื“_ื”ืฉื ื™ ืขื @guzyo ื”ืขืจื‘ ื‘-23:30 ื‘ื›ืืŸ 11 ื•ื‘ื›ืืŸ BOX
5
13
54
9,844
Starting in 1 HOUR at #SIGGRAPHAsia2025! Join us in Room S425 at 11:00 AM for our course: Master LoRA Training for Text-to-Video Models. Weโ€™re showing how to train a LoRA in just 1 hour using our open-source tools at @Lightricks. ๐Ÿ“ Room S425 โฐ 11:00 AM HKT (Today!) See you!
2
2
6
1,307
94
ื”ื’ืขื ื• ื ืฆื™ื’ื•ืช ืžื›ื•ื‘ื“ืช ืž @Lightricks ื‘ื›ื ืก ื”ืžืจืฉื™ื ืฉืœ GENML 2025 ื’ื ื‘ืชื•ืจ ื“ื•ื‘ืจื™ื ื’ื ื‘ืชื•ืจ ื ื•ืชื ื™ ื—ืกื•ืช, ื•ื‘ืขื™ืงืจ ื‘ืชื•ืจ ืžื•ื‘ื™ืœื™ื ื‘ืชื—ื•ื ื”ืขื‘ืจืชื™ ืกื“ื ื ืขืœ ืื™ืš ืœื”ืชืื™ื ืืช ื”ืžื•ื“ืœ ืฉืฉื™ื—ืจืจื ื• ืœืขื•ืœื ืœืฆืจื›ื™ื ื”ืกืคืฆื™ืคื™ื™ื ืฉืœ ื›ืœ ืื—ื“. ื”ื™ื” ืžืขื ื™ื™ืŸ ื•ื›ื™ืฃ ื’ื“ื•ืœ!
2
85
ื”ืคืขืœื” ืœืžืฆื‘ื™ืขื™ ื™ืžื™ืŸ: ื”ื ื” ืขื™ืงืจื™ ื—ื•ืง ื”ื’ื™ื•ืก ืฉืœ ื‘ื™ืกืžื•ื˜ ื›ืคื™ ืฉืคื•ืจืกืžื•. ืชื ื• ืœื™ื™ืง โคืื ืืชื ื—ื•ืฉื‘ื™ื ืฉื–ื” ื—ื•ืง ื˜ื•ื‘, ื•ืฆื™ื•ืฅ ืžื—ื“ืฉ ๐Ÿ”ƒ ืื ืืชื ื—ื•ืฉื‘ื™ื ืฉื–ื” ื—ื•ืง ืจืข. 1. ื™ืขื“ื™ ื’ื™ื•ืก: 4800 ื—ืจื“ื™ื ื‘ืฉื ื” ื”ืจืืฉื•ื ื”, 5,760 ื‘ืฉื ื” ื”ืฉื ื™ื™ื” ื•-6,840 ื‘ืฉืœื™ืฉื™ืช. 2. ื‘ื—ื•ืจื™ื ื—ื™ืœื•ื ื™ื™ื ืฉืœืžื“ื• ืฉื ืชื™ื™ื ื‘ืžื•ืกื“ ื—ืจื“ื™ ื™ืกืคืจื• ื›ื—ืœืง ืžื”ืžื›ืกื”. 3. ืฉื™ืจื•ืช ืื–ืจื—ื™/ืœืื•ืžื™ ื•ืฉื™ืจื•ืช ื‘ื–ืง"ื ื™ื™ื—ืฉื‘ื• ื›ื—ืœืง ืžื”ืžื›ืกื”. 4. ื”ื—ื•ืง ืœื ืžืชื—ื™ื™ื‘ ืœื’ื™ื™ืก ื—ืจื“ื™ื ืฆืขื™ืจื™ื ื•ืœื ื›ื•ืœืœ ืฉื•ื ื”ื‘ื˜ื—ื” ืœืฉื™ืจื•ืช ื‘ื™ื—ื™ื“ื•ืช ืงืจื‘ื™ื•ืช. 5. ื”ื’ื‘ืœื•ืช ื›ืœื›ืœื™ื•ืช ืขืœ ืžื™ ืฉืœื ื™ืชื’ื™ื™ืก ื™ื—ื•ืœื• ืจืง ืžื”ืฉื ื” ื”ืฉื ื™ื™ื” ื•ืจืง ื‘ืžืงืจื” ืฉืœ ืื™-ืขืžื™ื“ื” ื‘ื™ืขื“ื™ื. 6. ื‘ืžื™ื“ื” ื•ืœื ื™ืขืžื“ื• ื‘ื™ืขื“ื™ ื”ื’ื™ื•ืก, ืชื”ื™ื™ื ื” ืกื ืงืฆื™ื•ืช ืคืœื™ืœื™ื•ืช (ืื™ืกื•ืจ ื™ืฆื™ืื” ืžืŸ ื”ืืจืฅ ื•ืฉืœื™ืœืช ืจืฉื™ื•ืŸ) ื•ืื™ืกื•ืจ ืงื‘ืœืช ืžืœื’ื•ืช ืœื™ืžื•ื“, ื•ืกื ืงืฆื™ื•ืช ื—ืœืงื™ื•ืช ืขืœ ืชืงืฆื™ื‘ื™ ื”ื™ืฉื™ื‘ื•ืช 7. ืœืคื™ ื—ืœืง ืžื”ืคืจืกื•ืžื™ื ืชื”ื™ื” ื•ืขื“ื” ืฉืชื›ืœื•ืœ ื’ื ืจื‘ื ื™ื ืžื—ื•ืฅ ืœืฆื”"ืœ ืฉืชื•ื›ืœ ืœืงื‘ื•ืข ืฉื”ืžืกื’ืจืช ืฉื”ืฆื‘ื ืžืืคืฉืจ ืœื ื˜ื•ื‘ื” ืžืกืคื™ืง, ื•ืชื•ื›ืœ ืœืขืฆื•ืจ ืืช ื—ื•ื‘ืช ื”ื’ื™ื•ืก
9
2
868
Naomi Ken Korem retweeted
ืื—ื•ื–ื™ ื”ื’ื™ื•ืก ืฉืœ ื—ืจื“ื™ื ืœืคื ื™ ืฉื ื•ืช ื”ืฉื‘ืขื™ื ื ืขื™ื ื‘ื™ืŸ 50 ืœ75 ืื—ื•ื–. ืื—ื•ื–ื™ ื”ืขื‘ื•ื“ื” ื ืขื• ืกื‘ื™ื‘ ื”80 ืื—ื•ื–. ื•ืื—ื•ื– ื”ืชืœืžื™ื“ื™ื ื”ื—ืจื“ื™ื ื‘ื‘ืชื™ ื”ืกืคืจ ื”ื™ื” 4.ย  ืžื” ืงืจื” ื‘ืฉื ื•ืช ื”ืฉื‘ืขื™ื? ืžื“ื™ื ืช ื™ืฉืจืืœ ื”ืงื™ืžื” ืืช ื—ื‘ืจืช ื”ืœื•ืžื“ื™ื ื”ื—ืจื“ื™ืช. ื–ื• ืฉืœื ืžืชื’ื™ื™ืกืช, ืœื ืขื•ื‘ื“ืช, ื•ืžืชื ื’ื“ืช ืœืชืจื•ืžื” ืœืžื“ื™ื ื” ื‘ืื•ืคืŸ ื’ื•ืจืฃ. ื”ื™ื ืฉืžื” ืขืœ ื›ืš ืœื ืžืขื˜ ื›ืกืฃ.
10
15
212
7,247