🎁⌚ ¡SORTEO! ⌚🎁 Este Apple Watch Series 11 puede ser para ti:
🔺 Síguenos
🔺 Haz RT a este post
🔺 Y comenta para qué lo usarías con el hashtag #VodafoneHappy
Happy cumple 6 años 🎉 Y por eso...tienes otra oportunidad en la App Mi Vodafone. Hoy y mañana 17/10 sortearemos 2 más cada día. Más info en bio.
🎁REGALAMOS una gorra + tote bag de Pirelli. ¿Las quieres?
1) Síguenos en X e Instagram
2) Haz RT a este tweet
Elegiremos al ganador cuando acabe el #ItalianGP de #F1
Why do we even need to learn how to mix motions? 🤔In our previous work (DualMDM), we found something very interesting 💡: when mixing motions throughout the denoising process, it actually works better to mix them differently at different steps.
Our initial hypothesis was that early in the denoising process, interpersonal dynamics (how people relate to each other) matter more 👫. Later on the denoising, interpersonal dynamics (how a person moves individually) become more important 💃. So, we proposed different mixing schedulers that adapt across steps 🧠📈.
This approach worked better than others, but it still had limitations. While some schedulers performed better on average, none of them worked perfectly across all motion combinations 🎭. Different motion pairs needed slightly different schedulers to perform well.
💥 That’s what motivated MixerMDM! We introduce the Mixer — a lightweight transformer module that predicts a mixing weight at every step of the denoising process, conditioned on the motions to be mixed and their textual descriptions 🌀🤖. This allows us to dynamically and consistently mix motions while maintaining strong individual controllability 💃🕺✨
📢 Introducing MixerMDM [#CVPR2025] — the first learnable model composition technique for combining pre-trained human motion diffusion models while preserving their core characteristics. 🟡➕🔵🟰🟢
MixerMDM enables consistent, fine-grained individual controllability in generating human-human interactions by:
🎛️ Mixer – A lightweight module that learns how to dynamically mix motions based on their specific characteristics and input conditions.
👩🏻🔬 Mixing – A model composition formula that combines motions dynamically through the denoising process, guided by Mixer predictions.
👮🏻♂️ Adversarial Training – A training strategy using adversarial losses, where predictions from the pre-trained models act as pseudo-ground truths.
🏅 Evaluation – A novel metric for measuring mixing quality, evaluating both individual and interaction alignment.
❤️ Huge shoutout to my amazing supervisors and co-authors: @Germs96@cpalmeroc@SergioEscalera_ and José García-Rodríguez
🌍 Project: https://t.co/X3o25KZFw7
📚 Paper: https://t.co/wtniqZlp9D
🧬 Code: https://t.co/3W9zDG1siv
💃 in2IN: Leveraging individual Information to Generate Human INteractions 👫
🎊 Accepted at #CVPR2024 Workshop #HuMoGen 🎉
👫in2IN: interaction + individual descriptions
👫+💃 DualMDM: interaction model + individual prior = ⬆ ind. diversity + ⬆️ ind. controllability
-----
In this paper we present:
👫in2IN: a novel diffusion model architecture that is not only conditioned on the overall interaction description but also on the descriptions of the individual motion performed by each interactant.
[🥇 SOTA on InterHuman dataset 🥇]
👫+💃 DualMDM: a new motion composition technique to further increase the individual diversity and control. By combining our in2IN interaction model with a single-person (individual) motion prior, we generate interactions with more diverse intra-personal dynamics.
Huge ❤️thanks and credit ❤️ to my supervisors in my first paper as PhD student: @Germs96, @cpalmeroc, @SergioEscalera_, and José García-Rodríguez
🌍Web: https://t.co/dndGXBnf9o
📕arXiv: https://t.co/TXQRo9yhjb
💾GitHub: https://t.co/9KAyGMrb32
SORTEÍN 🚨🚨🚨 GP ESPAÑA F1 🇪🇸
Entrada doble en zona VIP para contemplar la Nº33 en riguroso directo, yo estaré también por ahí.
Valorado en 2178€
Para participar:
1) Seguir a @nanosecso y @thegptrips
2) RT a este tweet
Jueves hasta las 12 se puede participar. Suerte! 🍀
¡ChatGPT ha invadido internet! 🌎
He creado un recurso con los mejores indicaciones en español para usar ✅
¡Y es GRATIS! 🎁
Para obtenerlo:
1. Dale like al tweet
2. Retwittea
3. Comenta "Gracias"
Después de esto te llegará tu link por DM 📩
(Debes de estar siguiéndome)
🧳 REGALAMOS una maleta de Pirelli. ¿La quieres?
1) Síguenos en Twitter e Instagram
2) Haz RT a este tweet
¡Elegiremos al ganador cuando acabe el #JapaneseGP 🇯🇵!
*Sólo para España
NUEVO!
█▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀█
█🔥 𝟭𝟬𝟬€ 𝗣𝗔𝗥𝗔 𝗔𝗠𝗔𝗭𝗢𝗡 🔥█
█▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄█
Van a ser para una de las personas que haga 🔃RT y nos siga.
Dale mucho❤️ por el poder del chollo!
El jueves ganador/a
🔥🔥🔥🔥🔥🔥🔥🔥🔥
█▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀█
█░ 𝟭𝟬𝟬€ 𝗣𝗔𝗥𝗔 𝗔𝗠𝗔𝗭𝗢𝗡 ░█
█▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄█
Van a ser para una de las personas que haga 🔃RT y nos siga. Dale ❤️
El viernes decimos ganador/a.
🔥🔥🔥🔥🔥🔥🔥🔥🔥
🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠
50 EUROS para gastar en amazon porque sí van a ser para una de las personas que haga RT 🔁 y siga a Chollometro.
Dale mucho amor ❤️
Lunes decimos ganador
🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠🟠
⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️
50 EUROS para gastar en amazon (España) van a ser para una de las personas que haga retweet 🔁 y siga a Chollometro.
El poder del chollo te pide mucho ❤️
El lunes decimos ganador!
⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️⭐️
🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥
SI EN 3 HORAS ESTE TWEET LLEGA A 10k RETWEETS REGALAMOS 2 CHEQUES PARA AMAZON DE 50 EUROS A DOS DE LAS PERSONAS QUE RETWITEEN Y NOS SIGAN.
Dale ❤️ por el poder del chollo
🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥