正在加载视频...

视频加载失败

返回首页

We’ve open-sourced a new version of StreamingT2V, an advanced autoregressive method for generating high-quality, long videos with rich motion dynamics. Thanks to the @PicsartAI team for their outstanding work in democratizing AI!

humphrey_shi's profile picture

Humphrey Shi

4,604 subscribers

67,987 次观看 • 1 年前 •via X (Twitter)

科学技术教育
LIVE

Anya Rossi• Live Now

Private livecam show

6 条评论

clem 🤗 的头像
clem 🤗1 年前

@PicsartAI Would be so cool to have models on or demos on

Humphrey Shi 的头像
Humphrey Shi1 年前

@PicsartAI

Search Truth 的头像
Search Truth1 年前

@PicsartAI what about CogVideoX? which is much closer to more complex modern video generation than SVD simple panning shots.

白日梦想猿 的头像
白日梦想猿1 年前

@_akhaliq @PicsartAI 感谢分享，这就去尝试！

Sergio Molina | AI 的头像
Sergio Molina | AI1 年前

@PicsartAI Wow, this is awesome! Thanks for sharing the new version of StreamingT2V. The @PicsartAI team is doing incredible work in making AI accessible to everyone. Can't wait to see the impact of this AI

PICORN.com 的头像
PICORN.com1 年前

@PicsartAI We are all just the creation of a hyper intelligent AI.

相关视频

We just released a first version of StreamingT2V for long video generation with naive T2V/I2V support at: Thanks to the open-source community for sharing their models & AK Hugging Face for generous demo GPU grant! Have fun:
0:45

We just released a first version of StreamingT2V for long video generation with naive T2V/I2V support at: Thanks to the open-source community for sharing their models & AK Hugging Face for generous demo GPU grant! Have fun:

Humphrey Shi

27,793 次观看 • 2 年前

New on Poe: Kling 2.0! This advanced video model sets new standards for motion quality and image-to-video capabilities. It excels at transforming images you create on Poe, including those generated with GPT-4o, into high-quality videos. (1/2)
0:20

New on Poe: Kling 2.0! This advanced video model sets new standards for motion quality and image-to-video capabilities. It excels at transforming images you create on Poe, including those generated with GPT-4o, into high-quality videos. (1/2)

Poe

15,345 次观看 • 1 年前

Researchers at Meta recently shared MAGNeT, a single non-autoregressive transformer model for text-to-music & text-to-sound generation capable of generating audio on-par with the quality of SOTA models — at 7x the speed. MAGNeT is open source as part of AudioCraft. Hear audio samples and get more details on this work ➡️
0:51

Researchers at Meta recently shared MAGNeT, a single non-autoregressive transformer model for text-to-music & text-to-sound generation capable of generating audio on-par with the quality of SOTA models — at 7x the speed. MAGNeT is open source as part of AudioCraft. Hear audio samples and get more details on this work ➡️

AI at Meta

122,001 次观看 • 2 年前

Introducing The Matrix --- a foundation world model for generating infinite-length, hyper-realistic videos with real-time, frame-level control: - Infinite-length video generation - 720p high-quality rendering - Real-time, frame-level control at 16 FPS - Generalization to real-world video control 🔗Blog: 📄Paper: 💻Code & Playable Demo: Coming soon! Key Innovation: A brand new technique called the shift-window denoise process model, enabling auto-regressive generation for diffusion and consistency models in real-time. Special thanks to project leader Ruili Feng and the entire Matrix team for their dedication and hard work over the year-long project.
1:00

Introducing The Matrix --- a foundation world model for generating infinite-length, hyper-realistic videos with real-time, frame-level control: - Infinite-length video generation - 720p high-quality rendering - Real-time, frame-level control at 16 FPS - Generalization to real-world video control 🔗Blog: 📄Paper: 💻Code & Playable Demo: Coming soon! Key Innovation: A brand new technique called the shift-window denoise process model, enabling auto-regressive generation for diffusion and consistency models in real-time. Special thanks to project leader Ruili Feng and the entire Matrix team for their dedication and hard work over the year-long project.

Hongyang Zhang

178,322 次观看 • 1 年前

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:
1:39

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

Tencent Hy

89,257 次观看 • 10 个月前

Congrats to the Motion team on their $60M in funding at a $550M valuation! Motion is building the first agentic work suite for SMBs. Imagine the Microsoft suite, if it were built today with AI Agents natively embedded.
1:07

Congrats to the Motion team on their $60M in funding at a $550M valuation! Motion is building the first agentic work suite for SMBs. Imagine the Microsoft suite, if it were built today with AI Agents natively embedded.

Y Combinator

37,636 次观看 • 10 个月前

Excited to share that I’ve joined as a Motion Designer! A huge thank you to Ammaar Reshi , Jack, and the entire Eleven Labs team for the warm welcome and for believing in my work. Grateful for the opportunity to be part of such an innovative company pushing the boundaries of AI and creativity. Can’t wait to bring stories to life with motion!
0:26

Excited to share that I’ve joined as a Motion Designer! A huge thank you to Ammaar Reshi , Jack, and the entire Eleven Labs team for the warm welcome and for believing in my work. Grateful for the opportunity to be part of such an innovative company pushing the boundaries of AI and creativity. Can’t wait to bring stories to life with motion!

Aditya Verma

52,132 次观看 • 1 年前

An AVATAR program developed with UE5 that supports connecting to AI motion capture for real-time, high-quality UE animation output. It features a highly flexible character customization system and rich clothing and accessory assets. You can upload images to generate 3D models and quickly obtain high-quality 3D model assets.
1:52

An AVATAR program developed with UE5 that supports connecting to AI motion capture for real-time, high-quality UE animation output. It features a highly flexible character customization system and rich clothing and accessory assets. You can upload images to generate 3D models and quickly obtain high-quality 3D model assets.

CYANPUPPETS

11,542 次观看 • 5 个月前

Commendable work by the Bareilly Police team in creating a dazzling display of devotion and duty! Their dedication to ensuring a safe and joyous Diwali for all is truly inspiring. In recognition of their outstanding efforts, I am pleased to announce a reward for the team. Keep up the great work! #UPPolice
1:03

Commendable work by the Bareilly Police team in creating a dazzling display of devotion and duty! Their dedication to ensuring a safe and joyous Diwali for all is truly inspiring. In recognition of their outstanding efforts, I am pleased to announce a reward for the team. Keep up the great work! #UPPolice

DGP UP

72,980 次观看 • 1 年前

We’ve been looking for an excuse to work with GUNSHIP for too long, and couldn’t be more excited about this dark new version of Afterlife. Being a part of the Devil May Cry series has been an honor and too much fun. See you at the Game Awards- TONIGHT!
0:23

We’ve been looking for an excuse to work with GUNSHIP for too long, and couldn’t be more excited about this dark new version of Afterlife. Being a part of the Devil May Cry series has been an honor and too much fun. See you at the Game Awards- TONIGHT!

Evanescence

33,834 次观看 • 7 个月前

Over 5,000 new school places being delivered, including for children with SEND. With 95% of pupils getting their first choice, it’s a testament to the outstanding work of a dedicated and effective team. Proud to stand up for families and support real results.
0:43

Over 5,000 new school places being delivered, including for children with SEND. With 95% of pupils getting their first choice, it’s a testament to the outstanding work of a dedicated and effective team. Proud to stand up for families and support real results.

Cllr Charles Pugsley

49,204 次观看 • 1 年前

using MOCAP for player animation in UE5 requires "Generating Root Motion" that can drive the actual player character's locomotion during a runtime cinematic thanks Greg Richardson for the help with this demo in UE 5.6 with a new MetaHuman
0:59

using MOCAP for player animation in UE5 requires "Generating Root Motion" that can drive the actual player character's locomotion during a runtime cinematic thanks Greg Richardson for the help with this demo in UE 5.6 with a new MetaHuman

Matt Workman

13,070 次观看 • 1 年前

#SIGGRAPH2024 🥳🥳I've open-sourced the code for our Vertex Block Descent (VBD) paper as a part of Gaia: This is a collaborative work with Ziheng Liu. Thanks for your interest in VBD and Gaia! I released it ahead of schedule for you!
0:27

#SIGGRAPH2024 🥳🥳I've open-sourced the code for our Vertex Block Descent (VBD) paper as a part of Gaia: This is a collaborative work with Ziheng Liu. Thanks for your interest in VBD and Gaia! I released it ahead of schedule for you!

Dr. Anka He Chen

29,396 次观看 • 2 年前

Here's an edited version of the space by @jason with Elon Musk Bill Ackman and 2024 Democratic Presidential Candidate Dean Phillips. This version reduces time by 30% and significantly boosts audio quality. Link to original in replies. Thanks David Carbutt for the work!
46:11

Here's an edited version of the space by @jason with Elon Musk Bill Ackman and 2024 Democratic Presidential Candidate Dean Phillips. This version reduces time by 30% and significantly boosts audio quality. Link to original in replies. Thanks David Carbutt for the work!

Farzad 🇺🇸 🇮🇷

8,334,990 次观看 • 2 年前

Thanks to the amazing support for Ghostty scripting with ApplScript, I made a tiny skill for myself that can delegate work to new Anthropic Claude sessions in individual Ghostty splits 😍🤩 just what I wanted for so long. 🚀 Thanks Mitchell and team! 🔥
0:34

Thanks to the amazing support for Ghostty scripting with ApplScript, I made a tiny skill for myself that can delegate work to new Anthropic Claude sessions in individual Ghostty splits 😍🤩 just what I wanted for so long. 🚀 Thanks Mitchell and team! 🔥

Shai Mishali 🇮🇱

51,825 次观看 • 4 个月前

People in construction are now using AI to fake completed work, snapping a photo of an unfinished area, generating a finished version, and sending it to the builder to get paid. Diabolical.
1:36

People in construction are now using AI to fake completed work, snapping a photo of an unfinished area, generating a finished version, and sending it to the builder to get paid. Diabolical.

Financial Dystopia

1,512,352 次观看 • 6 个月前

Got five papers accepted by #ECCV2024 European Conference on Computer Vision #ECCV2026 ! Huge thanks to all my collaborators! 😃 See you in Milan 🇮🇹 Summary of Selected Works (I made a fast-forward for them 😄) - [Shape Generation] Surf-D: Generating High-Quality Surfaces of Arbitrary Topologies Using Diffusion Models, ECCV 2024. - [Efficient Motion Generation] EMDM: Efficient Motion Diffusion Model for Fast, High-Quality Human Motion Generation, ECCV 2024. - [Controllable Motion Generation] TLControl: Trajectory and Language Control for Human Motion Synthesis, ECCV 2024. - [Avatar Generation] Disentangled Clothed Avatar Generation from Text Descriptions, ECCV 2024. Project Page: Surf-D: EMDM: TLControl: SOSMPL:
3:49

Got five papers accepted by #ECCV2024 European Conference on Computer Vision #ECCV2026 ! Huge thanks to all my collaborators! 😃 See you in Milan 🇮🇹 Summary of Selected Works (I made a fast-forward for them 😄) - [Shape Generation] Surf-D: Generating High-Quality Surfaces of Arbitrary Topologies Using Diffusion Models, ECCV 2024. - [Efficient Motion Generation] EMDM: Efficient Motion Diffusion Model for Fast, High-Quality Human Motion Generation, ECCV 2024. - [Controllable Motion Generation] TLControl: Trajectory and Language Control for Human Motion Synthesis, ECCV 2024. - [Avatar Generation] Disentangled Clothed Avatar Generation from Text Descriptions, ECCV 2024. Project Page: Surf-D: EMDM: TLControl: SOSMPL:

Zhiyang (Frank) Dou

18,182 次观看 • 2 年前