Somdeb Sarkhel — accepted AI-conference papers

2026

GT-SVJ: Generative-Transformer-Based Self-Supervised Video Judge For Efficient Video Reward Modeling

Shivanshu Shekhar, Uttaran Bhattacharya, Raghavendra Addanki, Mehrab Tanjim, Somdeb Sarkhel, Tong Zhang

CVPR 2026

Aligning video generative models with human preferences remains challenging: current approaches rely on Vision-Language Models (VLMs) for reward modeling, but these models struggle to capture subtle temporal dynamics. We propose a fundamentally different approach: repurposing video generative models

Cited by 0 Source ↗Scholar ↗

2026

Mosaic: Runtime-Efficient Multi-Agent Embodied Planning

Kunjal Panchal, Saayan Mitra, Sunav Choudhary, Victor Bursztyn, Somdeb Sarkhel, Hui Guan

ICML 2026poster

LLM-based multi-agent embodied planning remains impractical due to prohibitively high execution latency. We identify failed actions as the dominant bottleneck, stemming from two core challenges: inaccurate state tracking under partial observability and inefficient coordination that produces redundan…

Cited by 0 Source ↗Scholar ↗

2025

Reparameterizing Hybrid Markov Logic Networks to handle Covariate-Shift in Representations

Anup Shakya, Abisha Thapa Magar, Somdeb Sarkhel, Deepak Venugopal

UAI 2025

We utilize Hybrid Markov Logic Networks (HMLNs) to combine embeddings learned from a Deep Neural Network (DNN) with symbolic relational knowledge. Since a DNN may not always learn optimal embeddings, we develop a mixture model to reduce variance in the HMLN parameterization. Further, we perform infe

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

SKALD: Learning-Based Shot Assembly for Coherent Multi-Shot Video Creation

Chen-Yi Lu, Md Mehrab Tanjim, Ishita Dasgupta, Somdeb Sarkhel, Gang Wu, Saayan Mitra +1

ICCV 2025poster

We present SKALD, a multi-shot video assembly method that constructs coherent video sequences from candidate shots with minimal reliance on text. Central to our approach is the Learned Clip Assembly (LCA) score, a learning-based metric that measures temporal and semantic relationships between shots…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

VISIAR: Empower MLLM for Visual Story Ideation

Zhaoyang Xia, Somdeb Sarkhel, Mehrab Tanjim, Stefano Petrangeli, Ishita Dasgupta, Yuxiao Chen +4

ACL 2025finding

Ideation, the process of forming ideas from concepts, is a big part of the content creation process. However, the noble goal of helping visual content creators by suggesting meaningful sequences of visual assets from a limited collection is challenging. It requires a nuanced understanding of visual…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2024

TAME-RD: Text Assisted Replication of Image Multi-Adjustments for Reverse Designing

Pooja Guhan, Uttaran Bhattacharya, Somdeb Sarkhel, Vahid Azizi, Xiang Chen, Saayan Mitra +2

ACL 2024findings

Given a source and its edited version performed based on human instructions in natural language, how do we extract the underlying edit operations, to automatically replicate similar edits on other images? This is the problem of reverse designing, and we present TAME-RD, a model to solve this problem…

Cited by 0 Source ↗PDF ↗Scholar ↗

2024

ToolChain: Efficient Action Space Navigation in Large Language Models with A Search

Yuchen Zhuang, Xiang Chen, Tong Yu, Saayan Mitra, Victor Bursztyn, Ryan A. Rossi +2

ICLR 2024poster

Large language models (LLMs) have demonstrated powerful decision-making and planning capabilities in solving complicated real-world problems. LLM-based autonomous agents can interact with diverse tools (e.g., functional APIs) and generate solution plans that execute a series of API function calls in…

Cited by 52 Source ↗PDF ↗Scholar ↗

2023

Flash: Concept Drift Adaptation in Federated Learning

Kunjal Panchal, Sunav Choudhary, Subrata Mitra, Koyel Mukherjee, Somdeb Sarkhel, Saayan Mitra +1

ICML 2023poster

In Federated Learning (FL), adaptive optimization is an effective approach to addressing the statistical heterogeneity issue but cannot adapt quickly to concept drifts. In this work, we propose a novel adaptive optimizer called Flash that simultaneously addresses both statistical heterogeneity and t…

Cited by 14 Source ↗PDF ↗Scholar ↗

2019

Adaptive Rao-Blackwellisation in Gibbs Sampling for Probabilistic Graphical Models

Craig Kelly, Somdeb Sarkhel, Deepak Venugopal

AISTATS 2019poster

Rao-Blackwellisation is a technique that provably improves the performance of Gibbs sampling by summing-out variables from the PGM. However, collapsing variables is computationally expensive, since it changes the PGM structure introducing factors whose size is dependent upon the Markov blanket of th…

Cited by 3 Source ↗PDF ↗code ↗Scholar ↗

2018

Efficient Weight Learning in High-Dimensional Untied MLNs

Khan Mohammad Al Farabi, Somdeb Sarkhel, Deepak Venugopal

AISTATS 2018poster

Existing techniques for improving scalability of weight learning in Markov Logic Networks (MLNs) are typically effective when the parameters of the MLN are tied, i.e., several ground formulas in the MLN share the same weight. However, to improve accuracy in real-world problems, we typically need to…

Cited by 0 Source ↗PDF ↗Scholar ↗

2015

Fast Lifted MAP Inference via Partitioning

Somdeb Sarkhel, Parag Singla, Vibhav G Gogate

NeurIPS 2015poster

Recently, there has been growing interest in lifting MAP inference algorithms for Markov logic networks (MLNs). A key advantage of these lifted algorithms is that they have much smaller computational complexity than propositional algorithms when symmetries are present in the MLN and these symmetries…

Cited by 20 Source ↗PDF ↗Scholar ↗