esha

founder @ hooked (

[esha @studioxesha]

be kind. keep it weird. lean into joy.

[esha @studioxesha]

@framer_x this is awesome! thanks for sharing your process!

[esha @studioxesha]

@EHuanglu lol what is sleep?🤪🫠

[esha @studioxesha]

@EHuanglu or i guess what *feels* like real time relative to any other progress that i've seen happen in my life!

[esha @studioxesha]

@EHuanglu yeah it's insane to be experiencing it get better in real time!

[esha @studioxesha]

@anuaakash 👀 this is great

[esha @studioxesha]

@EHuanglu sort of sounds like a disney princess voice in a mystery thriller movie, lol - very believable!

[21 @twentyone21___]

EXPLORERS

[Luma @LumaLabsAI]

👋Hello from @monster_library. Welcome to Monster Week!🎉 For the next 7 days, we're taking you on a journey through creativity, imagination, & pure magic ✨. Join us as we dive deep into world-building and storytelling, leading up to an exciting new short story reveal this weekend created by you! 📖🚀

[esha @studioxesha]

@Magiermogul @runwayml this is so good!!

[Particle Panic @Magiermogul]

Made it just in for the madness that is #GEN48 by @runwayml: 3,000 teams, 48 hours, 300,000 credits, drama, no sleep, creativity overdrive. Like a chef creating a dish and asking the ingredients what they want to become. Watch my entry "Breathkeeper" here: https://t.co/8BBmu4CSue

[𝙢𝙖𝙧𝙞𝙖_✨ @visual_dose]

Gm ✨
First day of school…📓

[esha @studioxesha]

@trbdrk this is awesome

[Linus ✦ Ekenstam @LinusEkenstam]

Best use of gen AI video/audio I’ve seen

Amazing work by: @trbdrk

[Gizem Akdag @gizakdag]

Do I need to say how much fun I have with what I do?

Midjourney + Luma + Suno

[AK @_akhaliq]

Loopy

Taming Audio-Driven Portrait Avatar with Long-Term Motion Dependency

paper page: https://t.co/9ovRUBHCCk

With the introduction of diffusion-based video generation techniques, audio-conditioned human video generation has recently achieved significant breakthroughs in both the naturalness of motion and the synthesis of portrait details. Due to the limited control of audio signals in driving human motion, existing methods often add auxiliary spatial signals to stabilize movements, which may compromise the naturalness and freedom of motion. In this paper, we propose an end-to-end audio-only conditioned video diffusion model named Loopy. Specifically, we designed an inter- and intra-clip temporal module and an audio-to-latents module, enabling the model to leverage long-term motion information from the data to learn natural motion patterns and improving audio-portrait movement correlation. This method removes the need for manually specified spatial motion templates used in existing methods to constrain motion during inference. Extensive experiments show that Loopy outperforms recent audio-driven portrait diffusion models, delivering more lifelike and high-quality results across various scenarios.

[esha @studioxesha]

@dustinhollywood @chariskm This is so beautiful 🥹

[esha @studioxesha]

@machine_mythos I'd love to chat with you about working on a project together. Please let me know the best way to message you?

[esha @studioxesha]

@machine_mythos this is impressive!

[esha @studioxesha]

@alillian This is so beautiful! I love your work so much. I'd love to chat with you about working on a project together, but I can't find a way to message you!

[esha @studioxesha]

@juliewdesign_ @FellowshipAi LOVE THIS