Hoppa till huvudinnehållet
Till KTH:s startsida

Kiran Chhatre

Profilbild av Kiran Chhatre

Doktorand

Detaljer

Adress
LINDSTEDTSVÄGEN 5
Rum

Om mig

Computer Vision for 3D Digital Humans

(Updated December 2024)

My research addresses the challenges of designing new techniques to create and control interactive virtual characters. I am working towards developing deep learning models enabling characters to exhibit appropriate behaviors in social interaction settings involving small groups, including adaptation to the style and manner of the interlocutor to support the smoothness of interaction and 'social bonding'. I am affiliated with Embodied Social Agents Lab and supervised by  Christopher Peters and Jonas Beskow. I am funded by the EU MSCA ITN CLIPE project.

I have completed a master's degree from RWTH Aachen University, Germany, and a bachelor's degree from the COEP Technological University, Pune India.

News

🚩 We present SPECTRUM: Open Vocabulary Human Parsing with Texture-based Diffusion Models [arXiv coming soon]

🚩 We present Audiopedia: Audio QA with Knowledge [arXiv coming soon]

🚩 Ph.D. internship with Electronic Arts (October 2024 - March 2025)

🚩 I participate in ICVSS 2024 in Sicily, Italy (7 - 13 July, 2024). During the event, I present a poster: AMUSE: Emotional Speech-driven 3D Body Animation via Disentangled Latent Diffusion (CVPR 2024).

🚩 I present a talk on "Virtual Characters in Small Conversational Groups" hosted by Prof Rajiv Soundararajan at Indian Institute of Science, Bangalore. (14 June 2024)

🚩Program Committee of CLIPE workshop @ Eurographics 2024, Limassol Cyprus. I present a talk on "Adaptive communicative social behaviors for virtual characters in small conversational groups".

🚩Research Ph.D. Intern with Adobe Research (April - July 2024)

🚩We present AMUSE: Emotional Speech-driven 3D Body Animation via Disentangled Latent Diffusion @ CVPR 2024 [Paper, Web, Code, Poster]

🚩 We present EMOTE: Emotional Speech-Driven Animation with Content-Emotion Disentanglement @ SIGGRAPH ASIA 2023 [Paper, Web, Code]

🚩 I present a talk on "Virtual Characters in Small Conversational Groups" hosted by Prof Carol O'Sullivan at Trinity College Dublin. (22 February 2024)

🚩Guest scientist with MPI Perceiving Systems, Tübingen (September 2022 - December 2024)

🚩 We present BEAMBayesOpt: Parallel Bayesian Optimization of Agent-based Transportation Simulation @ LOD 2022/ LNCS [Paper, Code, Wiki, Special Mentions]

🚩 Ph.D. internship with Ubisoft La Forge, Bordeaux (March - June 2022).

🚩 We present Modeling Groups of Conversational Agents and Tools for Virtual Crowds and Characters @ Summer School on Virtual Humans, Cyprus.

🚩We present Spatio-temporal priors in 3D human motion @ IEEE ICDL StEPP 2021 workshop.

🚩We present TEDxKTH Salon: AI – what should we focus on?

🚩Our Medium blog on Representation learning — The core of machine learning.

🚩We present the EU report on Social Group Formation in Virtual Environments.

🚩Our podcast with KTH AI Society on Natural Language Processing.

🚩Our webinar on Contrastive Learning for Visual Representations.

🚩Our webinar on Data-driven Motion Generation.

Academic Service

Teaching Assistant:
DH2323, DD2380, DD2421, DD2437DM1590.

Reviewer for:
IVA 2021, AAMAS 2021, IEEE COG 2021, LOD 2022AAMAS 2023, LOD 2023, AAMAS 2024, LOD 2024, ISMAR 2024, TVCG 2024, CVPR 2024, SIGGRAPH ASIA 2024, CHI 2025ICLR 2025.

Program committee for:
CLIPE workshop at Eurographics 2024

Industry Experience

I have previously worked with Lawrence Berkeley National Lab (USA), IBM Research (Germany), Dassault Systemes (Germany and India), APS - European Center for Mechatronics GmbH (Germany), RoboCup Logistics League (Germany)Qatar University, Tata Steel (India), Autocop Telematics (India), and Alfa Laval (India) on various projects.

Extracurricular @ KTH


Kurser

Artificiell intelligens (DD2380), assistent | Kurswebb

Artificiella neuronnät och djupa arkitekturer (DD2437), assistent | Kurswebb

Datorgrafik med interaktion (DH2323), assistent | Kurswebb