Fine-grained timbre and tempo control in video-to-audio synthesis via dual encoders. Disentangle style attributes without sacrificing semantic alignment. #Video2Audio #AudioGeneration #Timbre arxiv.org/abs/2606.07182v1
9
Audio Imitator: Controlling Timbre and Tempo in Video2Audio Synthesis with Audio Reference
Jiahui Zhao, Tianrui Wang, Chunyu Qiang, Cheng Gong, Xijuan Zeng, Feng Deng, Longbiao Wang
arxiv.org/abs/2606.07182 [𝚎𝚎𝚜𝚜.𝙰𝚂]
128
Jun 8
Jiahui Zhao, Tianrui Wang, Chunyu Qiang, Cheng Gong, Xijuan Zeng, Feng Deng, Longbiao Wang, "Audio Imitator: Controlling Timbre and Tempo in Video2Audio Synthesis with Audio Reference," arxiv.org/abs/2606.07182
359
Jan 17
既存のLTX-2 workflowのパラメータの見直し、新たなworkflow (Multi-frame I2V / video2audio / Temporal inpainting) の追加をしました
comfyui.nomadoor.net/ja/basi…
これで完成と思うたびに新たな情報が出てきたりして中々更新できませんでした…スイマセン(； ･`ω･´)
楽しんでいただければ！！
1
3
78
7,447
9 Sep 2025
Kijai has decided to update MMAudio. Its an older video2audio model in the same vein as HunYuan Foley.
I still want to go take a look at HunYuan Foley again, I feel like that model is probably better than MMAudio, although MMAudio might be faster.
github.com/kijai/ComfyUI-MMA…
2
147
🇮🇩 Kali Besar in Batavia in the Dutch East Indies in 1938
(current day Jakarta, Indonesia)
Input: original photo
-> img2img with Nano Banana "restore old photo"
-> then img2video with Kling 2.1 Master with ChatGPT-generated video prompt
-> then video2audio with MMAudio for background sounds
I think we'll have years of fun being able to finally interpret somewhat accurately the imperfect drawings people in human history made
And show how it actually would have looked if we had cameras then!
This is nano-banana with map input prompt: "full colour photograph. New Amsterdam in 1660. make sure it's full modern colors as if it's a photograph taken today."
This is old Dutch fortress of New Amsterdam, in 2025 the place in New York City where the National Museum of the American Indian is built on exactly that location!
16
10
136
69,850
LIVE playing PIRATE YAKUZA and shooting the shit about BONEPAINT BONECOIN BONEDUPS and VIDEO2AUDIO x.com/i/broadcasts/1vAGRDljP…
2
1
8
589
Used Video-to-Audio on the left clip to make her kitty purr. I am adding this to bonedub
We recently added MICROPHONE to bonedub so you can sing on your videos or do voices. Will do official announcement of the Microphone tool soon after some bug fixes.
Checkout this Video2Audio generation. This will be free to people making bonepaints.
4
3
19
961
4 Mar 2025
Looking for 1 intern on audio-visual generation (potentially video2audio generation)! We have the largest computation resources in Japan, and we do serious industrial research (and development). DM if interested, and you can find more about me in my homepage.
2
7
25
5,955
24 Feb 2025
Say hello to Video2Audio in #DreamMachine 🚀 @LumaLabsAI @gravicle thanks for adding magic to storytelling🚀Your innovation is transforming the way we create🎬
🔊One-Click Sound Generation
🎼Customizable Prompt
🔥Beta Access for All Users
Hit the "Audio" button & create magic!🌟
4
5
25
1,068
24 Feb 2025
💥💥VIDEO2AUDIO LIVE FOR FREE IN LUMA💥💥
For those who have an addiction to adding MMAUDIO to everything they possibly can this works along the lines of that but in my testing over the weekend this delivers a generally noticably higher level of audio quality.
Free while in Beta. So get those awesome crash landings going!
@LumaLabsAI #dreammachine #Ray2
24 Feb 2025
Video to Audio is now here in #DreamMachine. To generate sound for your video generations, just select the new "Audio" button. Create with a single click or describe with prompts for more customized direction. Audio is available now in beta for free to all users.
1
3
702
9 Feb 2025
for anyone wondering, that was me testing a video2audio model in @ComfyUI (sdxl, kling, mmaudio)
github.com/hkchengrex/MMAudi…
2
50
13 Jan 2025
Great 👍 ai is so advanced now; sora; kling and minimax; i hope if @Hailuo_AI add tools like mmaudio ( video2audio )
1
2
172
21 Dec 2024
Diffusion models are very strong and robust feature extractors, but recent works were only using them for recognition tasks. In our recent work (led by @MoayedHaji), we harness them for video2audio generation: they by far outperform conventional video feature extractors for audio/video temporal alignment and allow to achieve SotA results in sound quality as well: snap-research.github.io/AVLi…
20 Dec 2024
1/6 Introducing AV-Link, an approach to connect video and audio diffusion models in a self-contained framework to enable video-to-audio and audio-to-video generation with superb audio-video synchronization.
Project page: snap-research.github.io/AVLi…
1
13
1,376
20 Dec 2024
Say hello to a new semantically- and temporally-aligned #video2audio synthesis framework! 🎥🎶🎮
20 Dec 2024
🌟 Excited to Share Our Latest Work! 🎥🎶
Here we present Stable-V2A: Synthesis of Synchronized Sound Effects with Temporal and Semantic Controls
arxiv: arxiv.org/abs/2412.15023
web: ispamm.github.io/Stable-V2A
5
107
20 Dec 2024
Super interesting work on #GenAI #Video2Audio with impressive results from my friends @riccardofosco and @ChrisMarinoni with @EmilianPostola1 @marcomunita @cosmo_luca Joshua Reiss and @dhan90001!
👇 Go check it out!
20 Dec 2024
🌟 Excited to Share Our Latest Work! 🎥🎶
Here we present Stable-V2A: Synthesis of Synchronized Sound Effects with Temporal and Semantic Controls
arxiv: arxiv.org/abs/2412.15023
web: ispamm.github.io/Stable-V2A
1
10
328
15 Dec 2024
Chinese Giant Network public the "QianYing" voice game generation large model.
Including：video model YingGame and video2audio model YingSound
Another trillion $$$ industry (Gaming) is Kaput.
青龍聖者
1
3
1,155
