Posts by Collection
mlmonad
portfolio
publications
Genetic Bi-objective Optimization Approach to Habitability Score
Published in First International Conference on Modeling, Machine Learning and Astronomy, 2019
We show the applicability of genetic algorithms in optimizing the habitability of exoplanets.
Gestop - Customizable Gesture Control of Computer Systems
Published in Proceedings of the 3rd ACM India Joint International Conference on Data Science & Management of Data, 2021
This paper presents Gestop, a framework to interface with computer systems with hand gestures.
Searching a Raw Video Database using Natural Language Queries
Published in International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT), 2021
This work aims to provide an end-to-end pipeline to search a video database with a voice query from the end user.
Using Sampling to Estimate and Improve Performance of Automated Scoring Systems with Guarantees
Published in The 12th Symposium on Educational Advances in Artificial Intelligence (EAAI-AAAI22), 2021
We propose a sampling scheme to improve performance of Human-in-the-loop Automated Scoring Systems.
DeepSmooth - Efficient and Smooth Depth Completion
Published in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (VOCVALC Workshop), 2023
We propose a model that propagates information efficently for the task of depth completion, enforcing depth consistency spatio-temporally and producing smooth depth maps.
Beyond Words - A Topological Exploration of Coherence in Text Documents
Published in The Second Tiny Papers Track at ICLR, 2024
In this work, we show the use of Topological Data Analysis (TDA) techniques on attention maps of language models to model the coherence of text.
BG-HOP: A Bimanual Generative Hand-Object Prior
Published in Agents in Interaction, from Humans to Robots workshop at CVPR, 2025
In this work, we present BG-HOP, a generative prior that seeks to model bimanual hand-object interactions in 3D using limited bimanual interaction data and demonstrate preliminary results in capturing the joint distribution of hands and objects.
GHOST: Hierarchical Sub-Goal Policies for Generalizing Robot Manipulation
Published in Robotics: Science and Systems (RSS), 2026
We present GHOST, which factorizes manipulation into a high-level policy that predicts the next sub-goal as a distribution over 3D end-effector poses from multi-view RGB-D, and a low-level goal-conditioned controller that executes embodiment-specific actions, enabling generalization and cross-embodiment transfer.
From Grasps to Dexterity: Large-Scale Grasp Pretraining for Dexterous Manipulation
Published in Conference on Robot Learning (CoRL), 2026
We show that large-scale dexterous grasp datasets can do more than grasp synthesis: pretraining the low-level controller of a hierarchical policy on 355k grasp trajectories enables functional dexterity in articulated tool use, which we evaluate on DexCraft, a new benchmark of six tool-use tasks.
H2RBench: A Real-to-Sim Benchmark for Evaluating Human-to-Robot Transfer
Published in Conference on Robot Learning (CoRL), 2026
We present H2RBench, a Real2Sim benchmark that evaluates human-to-robot transfer methods under a standardized protocol across four manipulation tasks, characterizing how each method scales with human demonstration data and showing that simulation performance is strongly predictive of real-world performance.
An Empirical Study on What Matters for Viewpoint-Generalizable Policies in Visual Imitation Learning
Published in arXiv preprint, 2026
We present a controlled empirical study of which design choices let visuomotor policies generalize across camera viewpoints, finding that retaining dense visual tokens and letting the action head participate in geometric reasoning yields a policy that transfers zero-shot from simulation to the real world under random camera configurations.
talks
This is a description of your talk, which is a markdown file that can be all markdown-ified like any other post. Yay markdown!