Media is too big
VIEW IN TELEGRAM
Meet another text-to-video AI from China - Minimax
Judging by this video, Minimax is not inferior in quality to its Kling sibling
You bet, because they have such investment giants as Alibaba and Tencent behind them.
The maximum length of the clip is limited to 6 seconds.
You can try it here 👈
#service #aitools #videogeneration #video
Brain Storm
Judging by this video, Minimax is not inferior in quality to its Kling sibling
You bet, because they have such investment giants as Alibaba and Tencent behind them.
The maximum length of the clip is limited to 6 seconds.
You can try it here 👈
#service #aitools #videogeneration #video
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
Sora's counterpart on your PC! 🦾
The Chinese have introduced the CogVideoX 5B model, which generates videos in decent quality.
It requires at least 12GB of video memory to run. The maximum video length is 6 seconds at a resolution of 720 x 480.
The smaller model, CogVideoX 2B, can even run on a 1080TI.
The model is available on GitHub
You can try it online in demo mode here
#news #video #videogeneration #aitools
Brain Storm
The Chinese have introduced the CogVideoX 5B model, which generates videos in decent quality.
It requires at least 12GB of video memory to run. The maximum video length is 6 seconds at a resolution of 720 x 480.
The smaller model, CogVideoX 2B, can even run on a 1080TI.
The model is available on GitHub
You can try it online in demo mode here
#news #video #videogeneration #aitools
Brain Storm
Powerful Text to Video Generator - Hotshot
- Hotshot Act-One is the best toolkit for creating three-second videos.
- Hotshot can make ten second videos in "ok" quality
It even has a simple prompting guide inside
#service #videogeneration #aitools #video
Brain Storm
- Hotshot Act-One is the best toolkit for creating three-second videos.
- Hotshot can make ten second videos in "ok" quality
It even has a simple prompting guide inside
#service #videogeneration #aitools #video
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
While we were asleep, Adobe introduced its video AI – Firefly Video 🎥
The quality is on par with Sora and surpasses Chinese AI models like Kling and Minimax. The realism is top-notch!
The video sequence can be generated both from a text prompt and an image.
Features like changing the viewing angle and camera movement have become commonplace. But Firefly still has surprises: check out the clip where a dandelion is examined through a magnifying glass. Can other AIs replicate that?
One of Firefly Video’s advantages is the ability to animate words.
There’s also a cool feature that lets you add generated effects (atmospheric elements like fire, smoke, dust particles, and water on a black or green background) to your original video directly in Premiere Pro and After Effects.
More details are available here, where you can also sign up for the waitlist.
#news #videogeneration #video
Brain Storm
The quality is on par with Sora and surpasses Chinese AI models like Kling and Minimax. The realism is top-notch!
The video sequence can be generated both from a text prompt and an image.
Features like changing the viewing angle and camera movement have become commonplace. But Firefly still has surprises: check out the clip where a dandelion is examined through a magnifying glass. Can other AIs replicate that?
One of Firefly Video’s advantages is the ability to animate words.
There’s also a cool feature that lets you add generated effects (atmospheric elements like fire, smoke, dust particles, and water on a black or green background) to your original video directly in Premiere Pro and After Effects.
More details are available here, where you can also sign up for the waitlist.
#news #videogeneration #video
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
A fresh AI for video deepfakes 👁🗨
It takes any source - your own videos or cuts from movies while understanding even the most complex references
Get it here
#service #aitools #videogeneration
Brain Storm
It takes any source - your own videos or cuts from movies while understanding even the most complex references
Get it here
#service #aitools #videogeneration
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
👁🗨 Comparison of #Gen3 and #Kling 1.5
I originally made the flamethrower commercial in Gen-3, so there is a way to visually compare the footage. Notice how much more movement Kling makes.
#video #videoai #videogeneration
Brain Storm
I originally made the flamethrower commercial in Gen-3, so there is a way to visually compare the footage. Notice how much more movement Kling makes.
#video #videoai #videogeneration
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
👁🗨 MIMO Neural Network Completely Transforms Characters in Videos
MIMO is a new video-to-video model from Alibaba. It can imitate anyone, anywhere, even during complex movements with object interactions.
With a reference image, MIMO can synthesize animated avatars in just a few minutes.
GitHub and neural network's main page
#Service #useful #video #videogeneration
Brain Storm
MIMO is a new video-to-video model from Alibaba. It can imitate anyone, anywhere, even during complex movements with object interactions.
With a reference image, MIMO can synthesize animated avatars in just a few minutes.
GitHub and neural network's main page
#Service #useful #video #videogeneration
Brain Storm
Media is too big
VIEW IN TELEGRAM
🔴 Circles in human civilization
An enthusiast made a video using the transition function from one frame to another. Made with Midjourney, Kling 1.0 and Luma. Not all the transitions turned out well but they may have generated the most dramatic egg cooking scene ever!
#fun #video #videogeneration
Brain Storm
An enthusiast made a video using the transition function from one frame to another. Made with Midjourney, Kling 1.0 and Luma. Not all the transitions turned out well but they may have generated the most dramatic egg cooking scene ever!
#fun #video #videogeneration
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
The ultimate neural network killer! 🔫
Bytedance has introduced Dreamina, a universal content generator.
Dreamina offers a wide range of features, including image and video generation, music creation, editing (storyboarding, motion brush, canvas tools, and much more).
Essentially, this neural network covers all the main creative functions one would expect from AI.
Here's the link, authorization requires a TikTok account.
As we predicted, the super-app is on the horizon.
#service #aitools #videogeneration
Brain Storm
Bytedance has introduced Dreamina, a universal content generator.
Dreamina offers a wide range of features, including image and video generation, music creation, editing (storyboarding, motion brush, canvas tools, and much more).
Essentially, this neural network covers all the main creative functions one would expect from AI.
Here's the link, authorization requires a TikTok account.
As we predicted, the super-app is on the horizon.
#service #aitools #videogeneration
Brain Storm
Runway has rolled out another awesome feature — advanced camera control 🎥
Now in Gen-3 Alpha Turbo, you can choose the direction and intensity of movement across the scene.
Rotate, pan, and control the camera however you want!
Another breakthrough from Runway! 🚀
#service #aitools #videogeneration #video
Brain Storm
Now in Gen-3 Alpha Turbo, you can choose the direction and intensity of movement across the scene.
Rotate, pan, and control the camera however you want!
Another breakthrough from Runway! 🚀
#service #aitools #videogeneration #video
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
👁🗨 ByteDance has announced X-Portrait 2
This AI brings photos to life with ULTRA-realistic expressions and emotions transferred from a single video. Lip sync with speech is included too. 😮
It works beautifully with both realistic portraits and animations.
Project page: It even includes a comparison with Runway Act-One, and X-Portrait 2 does a much better job at bringing portraits to life 🔥
#service #videogeneration #aitools
Brain Storm
This AI brings photos to life with ULTRA-realistic expressions and emotions transferred from a single video. Lip sync with speech is included too. 😮
It works beautifully with both realistic portraits and animations.
Project page: It even includes a comparison with Runway Act-One, and X-Portrait 2 does a much better job at bringing portraits to life 🔥
#service #videogeneration #aitools
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
HeyGen can now create ultra-realistic avatars from your photos 📸
Upload a few pictures, and HeyGen will generate your personalized model, producing hyper-realistic avatars with natural movements and lip sync. 👄
You can easily change outfits, backgrounds, and settings using prompts.
AI blogging is reaching a whole new level! 🚀
#service #aitools #videogeneration
Brain Storm
Upload a few pictures, and HeyGen will generate your personalized model, producing hyper-realistic avatars with natural movements and lip sync. 👄
You can easily change outfits, backgrounds, and settings using prompts.
AI blogging is reaching a whole new level! 🚀
#service #aitools #videogeneration
Brain Storm
Google DeepMind has unveiled CAT4D, an AI that transforms regular videos into immersive 4D experiences! 🎥✨
CAT4D uses a specialized model trained on vast amounts of diverse video data to recreate scenes from any angle and at any point in time.
Users can simply rotate the video in the desired direction to explore. 🌌🔄
The tool works with both generated and real-world videos, opening up exciting possibilities for creators and analysts alike.
Check out the project's GitHub for examples. 🚀
#AI #DeepMind #service #videogeneration #video
Brain Storm
CAT4D uses a specialized model trained on vast amounts of diverse video data to recreate scenes from any angle and at any point in time.
Users can simply rotate the video in the desired direction to explore. 🌌🔄
The tool works with both generated and real-world videos, opening up exciting possibilities for creators and analysts alike.
Check out the project's GitHub for examples. 🚀
#AI #DeepMind #service #videogeneration #video
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
BOOM: Pika has launched a new version of its generator! 💥🎥
The highlight feature? Video generation from your images, creating a seamless composition.
You can upload:
Your photo
Images of a location and outfits
And voilà—a video featuring your face is ready! 🖼️🎬
#AI #Pika #VideoGeneration #Innovation #service #aitools
Brain Storm
The highlight feature? Video generation from your images, creating a seamless composition.
You can upload:
Your photo
Images of a location and outfits
And voilà—a video featuring your face is ready! 🖼️🎬
#AI #Pika #VideoGeneration #Innovation #service #aitools
Brain Storm
This media is not supported in your browser
VIEW IN TELEGRAM
A new player on the AI scene: MinT AI 🎥✨
MinT AI is a groundbreaking video generator capable of creating multiple perspectives of the same object while allowing precise control over actions and emotions in each frame.
Example:
[0.0s → 0.02s]: A girl stands, head turned left, arms crossed.
[0.03s → 0.05s]: The girl looks at the camera, lowers her arms, and smiles.
[0.05s → 0.09s]: She raises her right hand and waves at the camera.
[0.09s → 0.12s]: She stands with her right hand down and her left hand resting at her waist.
This level of control ensures consistency in video storytelling—a game-changer for creators. 🚀
Full-length AI-generated films are getting closer!
Check out the project page for more details.
#AI #MinTAI #VideoGeneration #Innovation #aitools #service
Brain Storm
MinT AI is a groundbreaking video generator capable of creating multiple perspectives of the same object while allowing precise control over actions and emotions in each frame.
Example:
[0.0s → 0.02s]: A girl stands, head turned left, arms crossed.
[0.03s → 0.05s]: The girl looks at the camera, lowers her arms, and smiles.
[0.05s → 0.09s]: She raises her right hand and waves at the camera.
[0.09s → 0.12s]: She stands with her right hand down and her left hand resting at her waist.
This level of control ensures consistency in video storytelling—a game-changer for creators. 🚀
Full-length AI-generated films are getting closer!
Check out the project page for more details.
#AI #MinTAI #VideoGeneration #Innovation #aitools #service
Brain Storm
Google keeps catching up and rolling out solid AI tools 🚀
Today, they announced Veo 2 — a video generator competing with Sora, Kling, and others. 🎥
What Veo 2 can do:
Generate videos in 4K resolution.
Understand camera controls in prompts, like:
Wide shots.
First-person view.
Drone footage.
Accurately recreate real-world physics and human facial expressions.
But for now, these are just promises—the polished demos could be the result of hundreds of hours of cherry-picking. 🍒
Let’s wait for the release before jumping to conclusions.
Sign up for the waitlist and read more here 👈
#AI #Google #Veo2 #VideoGeneration #service #aitools
Brain Storm
Today, they announced Veo 2 — a video generator competing with Sora, Kling, and others. 🎥
What Veo 2 can do:
Generate videos in 4K resolution.
Understand camera controls in prompts, like:
Wide shots.
First-person view.
Drone footage.
Accurately recreate real-world physics and human facial expressions.
But for now, these are just promises—the polished demos could be the result of hundreds of hours of cherry-picking. 🍒
Let’s wait for the release before jumping to conclusions.
Sign up for the waitlist and read more here 👈
#AI #Google #Veo2 #VideoGeneration #service #aitools
Brain Storm
Kling 1.6 Update is Here! 🚀🎥
The latest version of Kling brings major improvements:
Enhanced prompt responsiveness: The AI now better understands and precisely follows your text prompts. ✍️✅
Upgraded visuals and physics: Improved aesthetic quality and more realistic physical actions.
A Shift in the Video-Gen Race:
The focus has moved from sheer quality to precise prompt adherence and realistic physics.
Currently leading the pack:
Veo 2
Minimax
Kling 1.6
The competition is heating up! 🔥
#AI #Kling #VideoGeneration #Update #service
The latest version of Kling brings major improvements:
Enhanced prompt responsiveness: The AI now better understands and precisely follows your text prompts. ✍️✅
Upgraded visuals and physics: Improved aesthetic quality and more realistic physical actions.
A Shift in the Video-Gen Race:
The focus has moved from sheer quality to precise prompt adherence and realistic physics.
Currently leading the pack:
Veo 2
Minimax
Kling 1.6
The competition is heating up! 🔥
#AI #Kling #VideoGeneration #Update #service
This media is not supported in your browser
VIEW IN TELEGRAM
A video generator battle has begun! 🎥🤖
The creator of the video compared the context understanding of the top video AI models.
For the test, they fed them the same prompt:
"A pair of hands skillfully slicing a perfectly cooked steak on a wooden cutting board. A faint steam rises from it."
This test is considered extremely challenging for AI models: hands, the physics of slicing, natural movements, interpreting "perfectly cooked steak," steam, juices, etc.
And as it turns out, the new Veo 2 from Google performed the best! 🔥
Minimax was a close second.
The rest, including Runway Gen3, generated some crazy results, especially with raw meat and knives sinking into it. 😅
Google really worked hard, breaking into the AI-video race! 🚀
#AI #Veo2 #VideoGeneration #Technology #Innovation
Brain Storm
The creator of the video compared the context understanding of the top video AI models.
For the test, they fed them the same prompt:
"A pair of hands skillfully slicing a perfectly cooked steak on a wooden cutting board. A faint steam rises from it."
This test is considered extremely challenging for AI models: hands, the physics of slicing, natural movements, interpreting "perfectly cooked steak," steam, juices, etc.
And as it turns out, the new Veo 2 from Google performed the best! 🔥
Minimax was a close second.
The rest, including Runway Gen3, generated some crazy results, especially with raw meat and knives sinking into it. 😅
Google really worked hard, breaking into the AI-video race! 🚀
#AI #Veo2 #VideoGeneration #Technology #Innovation
Brain Storm
ByteDance выкатила Bernini — новую систему для генерации и редактирования видео, где MLLM-модель сначала «планирует» сцену на семантическом уровне, а уже потом диффузионная модель рендерит итоговый ролик. По сути, ИИ сначала думает, потом рисует. 🤯
Проект построен на базе WAN 2.2 и дополнительных чекпоинтах от команды ByteDance, а также моделей, указанных авторами в репозитории. Поддерживаются генерация видео, редактирование, работа с референсами и вставка объектов в готовые ролики.
Потестить можно тут:
bernini-ai.github.io
#AI #VideoGeneration #ByteDance #WAN22 #pepewtfa
• Sent via TeleFeed
Проект построен на базе WAN 2.2 и дополнительных чекпоинтах от команды ByteDance, а также моделей, указанных авторами в репозитории. Поддерживаются генерация видео, редактирование, работа с референсами и вставка объектов в готовые ролики.
Потестить можно тут:
bernini-ai.github.io
#AI #VideoGeneration #ByteDance #WAN22 #pepewtfa
• Sent via TeleFeed
This media is not supported in your browser
VIEW IN TELEGRAM
Nvidia releases open-source video upscaler
🚀 Nvidia released Open-H3-Regenerate-2K-IC, an open-source architecture that upscales video to 4K.
The model uses in-context mechanisms to fix artifacts like face drift and text blur, keeping the original source draft within the transformer context during generation.
#ArtificialIntelligence ⚡ #Nvidia ⚡ #VideoGeneration
Subscribe
• Sent via TeleFeed
🚀 Nvidia released Open-H3-Regenerate-2K-IC, an open-source architecture that upscales video to 4K.
The model uses in-context mechanisms to fix artifacts like face drift and text blur, keeping the original source draft within the transformer context during generation.
#ArtificialIntelligence ⚡ #Nvidia ⚡ #VideoGeneration
Subscribe
• Sent via TeleFeed