BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260805T205503Z
LOCATION:Room 403 A
DTSTART;TZID=America/Los_Angeles:20260723T154500
DTEND;TZID=America/Los_Angeles:20260723T171500
UID:siggraph_SIGGRAPH 2026_sess101@linklings.com
SUMMARY:Multimodal Learning
DESCRIPTION:PointLLM-R: Enhancing 3D Point Cloud Reasoning via Chain-of-Th
 ought\n\nPointLLM-R introduces chain-of-thought reasoning for 3D point clo
 ud understanding via a data-centric pipeline. It builds a large reasoning 
 dataset (PoCoTI) and trains a multimodal model that achieves state-of-the-
 art performance, improving accuracy, interpretability, and robustness acro
 ss synthetic ...\n\n\nChaoqi Chen, Qile Xu, Wenjun Zhou, and Hui Huang (Sh
 enzhen University)\n---------------------\nReciprocal Latent Fields for Pr
 ecomputed Sound Propagation\n\nWave-coding methods for sound propagation i
 n games are very memory-intensive. We present Reciprocal Latent Fields lev
 eraging Riemannian metric learning to reduce memory usage by orders of mag
 nitude. Our method infers relevant acoustic parameters nearly instantaneou
 sly and results in perceptually ide...\n\n\nHugo Seuté (Ubisoft Divertisse
 ments Inc.) and Pranai Vasudev, Etienne Richan, and Louis-Xavier Buffoni (
 Audiokinetic Inc.)\n---------------------\nMAViD: A Multimodal Framework f
 or Audio-Visual Dialogue Understanding and Generation\n\nMAViD is a multim
 odal framework for interactive audio-visual dialogue. Its Conductor-Creato
 r architecture integrates multimodal understanding and reasoning into resp
 onse generation. By combining AR and diffusion models, MAViD enables synch
 ronized, long-duration audio-video generation with consistent...\n\n\nYoux
 in Pang (Tsinghua University); Jiajun Liu, Lingfeng Tan, Yong Zhang, and F
 eng Gao (meituan); Xiang Deng (Tsinghua University); Zhuoliang Kang and Xi
 aoming Wei (meituan); and Yebin Liu (Tsinghua University)\n---------------
 ------\nAudio-Omni: Extending Multi-modal Understanding to Versatile Audio
  Generation and Editing\n\nAudio-Omni is the first unified framework for m
 ulti-modal audio generation and editing across sound, music, and speech. B
 y combining a frozen multimodal language model with a trainable diffusion 
 transformer, it achieves state-of-the-art performance on multiple benchmar
 ks while exhibiting remarkable ...\n\n\nZeyue Tian (Hong Kong University o
 f Science and Technology), Binxin Yang (Tencent), Zhaoyang Liu (Hong Kong 
 University of Science and Technology), Jiexuan Zhang (Peking University), 
 Ruibin Yuan (Hong Kong University of Science and Technology), Hubery Yin (
 Tencent), Qifeng Chen (Hong Kong University of Science and Technology), Ch
 en Li and Jing Lyu (Tencent), and Wei Xue and Yike Guo (Hong Kong Universi
 ty of Science and Technology)\n---------------------\nJust-Dub-It: Video d
 ubbing via Joint Audio-Visual Diffusion\n\nJUST-DUB-IT is a unified audio-
 video diffusion model for automatic video dubbing. It jointly synthesizes 
 translated speech and synchronized lip movements via a lightweight LoRA ad
 apter, replacing traditional pipeline approaches. Trained on synthetically
  generated multilingual data with strategic inp...\n\n\nAnthony Chen (Peki
 ng University, Lightricks); Naomi Ken Korem, Tavi Halperin, Matan Ben Yose
 f, Urska Jelercic, and Ofir Bibi (Lightricks); and Or Patashnik and Daniel
  Cohen-Or (Tel Aviv University)\n---------------------\nVfxDB: A Visual Ef
 fects Volume Dataset and Benchmark for VDB-Native Generative Modeling\n\nV
 fxDB introduces a pioneering 1-million-sample dataset tailored for VDB-nat
 ive generation of sparse volumetric VFX sequences, alongside a reproducibl
 e benchmark suite. Validated by a scalable diffusion baseline featuring ou
 r Atomic-Continuous prior, this comprehensive pipeline establishes essenti
 al ...\n\n\nJunwei Shu (East China Normal University); Hantang Liu (KunByt
 e); Dawei Miao (Independent Researcher); and Wenzheng Song, Mingyang Yuan,
  Wenjie Liu, Changgu Chen, Yang Li, and Changbo Wang (East China Normal Un
 iversity)\n---------------------\nMultimodal Learning - Interactive Discus
 sion\n\nAfter the summary presentations, attendees will participate in an 
 interactive discussion. Outside the room will be a series of poster boards
  for authors to gather around with the audience. Authors are invited to br
 ing any material related to their paper that could instigate further conve
 rsation such...\n\n\nInterest Area: Research & Education\n\nKeyword: Anima
 tion, Artificial Intelligence/Machine Learning, Games, Generative AI, Geom
 etry, Modeling, Simulation\n\nRegistration Category: Full Conference Suppo
 rter, Full Conference\n\nSession Chair: Przemyslaw Musialski (New Jersey I
 nstitute of Technology (NJIT))
END:VEVENT
END:VCALENDAR