Publications

2026

1. paper

   Wiggle and Go! System Identification for Zero-Shot Dynamic Rope Manipulation

   Arthur Jakobsson, Abhinav Mahajan, Karthik Pullalarevu, and 6 more authors

   2026

   Accepted, CoRL 2026

   Abs

   Many robotic tasks are unforgiving; a single mistake in a dynamic throw can lead to unacceptable delays or unrecoverable failure. To mitigate this, we present a novel approach that leverages learned simulation priors to inform goal-conditioned dynamic manipulation of ropes for efficient and accurate task execution. Related methods for dynamic rope manipulation either require large real-world datasets to estimate rope behavior or the use of iterative improvements on attempts at the task for goal completion. We introduce Wiggle and Go!, a system-identification, two-stage framework that enables zero-shot task rope manipulation. The framework consists of a system identification module that observes rope movement to predict descriptive physical parameters, which then informs an optimization method for goal-conditioned action prediction for the robot to execute zero-shot in the real. Our method achieves strong performance across multiple dynamic manipulation tasks enabled by the same task-agnostic system identification module which offers seamless switching between different manipulation tasks, allowing a single model to support a diverse array of manipulation policies. We achieve a 3.55 cm average accuracy on 3D target striking in real using rope system parameters in comparison to 15.34 cm accuracy when our task model is not system-parameter-informed. We achieve a Pearson correlation coefficient of 0.95 between Fourier frequencies of the predicted and real ropes on an unseen trajectory.

2. paper

   MERIT: Mitigating Exposure Bias in Generative XMC for User-Interest Propensity Modelling

   Abhinav Mahajan, Arindam Sarkar, and Prakash Mandayam Comar

   2026

   Accepted, EMNLP 2026

   Abs

   Matching users to interest categories at scale is central to personalized shopping, but the task is challenging in large e-commerce platforms, where label spaces continually evolve and user-interest signals are sparse and long-tailed. Autoregressive language models are appealing because their world knowledge and semantic priors over descriptors generalize across extreme label spaces and accommodate multiple valid label assignments. Yet under teacher-forced fine-tuning, inference-time predictions become part of the conditioning context: early errors steer later outputs toward co-occurring labels, over-generating near-correlates and missing unrelated true interests. We present MERIT, a framework for user-interest propensity modeling that mitigates this exposure bias through a self-correction objective. A permutation-invariant multi-target loss over shuffled mixtures of gold and mined hard-negative labels exposes the generator to erroneous prefixes while preserving the efficiency of teacher-forced training. This training objective concentrates supervision at classification positions, yielding propensity-aligned hidden states powering a lightweight scorer for bidirectional retrieval (interests for users and users for interests). On a proprietary e-commerce dataset with 250k+ interest categories, MERIT improves global recall by at least 11.9% and average Hit@k by 6.1%. In production A/B tests, it achieves +0.26% gain in user conversion.

3. paper

   Towards Design Compositing

   Abhinav Mahajan, Abhikhya Tripathy, Sudeeksha Reddy Pala, and 3 more authors

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Jun 2026

   Abs

   Graphic design creation involves harmoniously assembling multimodal components such as images, text, logos, and other visual assets collected from diverse sources, into a visually-appealing and cohesive design. Recent methods have largely focused on layout prediction or complementary element generation, while retaining input elements exactly, implicitly assuming that provided components are already stylistically harmonious. In practice, inputs often come from disparate sources and exhibit visual mismatch, making this assumption limiting. We argue that identity-preserving stylization and compositing of input elements is a critical missing ingredient for truly harmonized components-to-design pipelines. To this end, we propose GIST, a training-free, identity-preserving image compositor that sits between layout prediction and typography generation, and can be plugged into any existing components-to-design or design-refining pipeline without modification. We demonstrate this by integrating GIST with two substantially different existing methods, LaDeCo and Design-o-meter. GIST shows significant improvements in visual harmony and aesthetic quality across both pipelines, as validated by LLaVA-OV and GPT-4V on aspect-wise ratings and pairwise preference over naive pasting.

2025

1. paper

   Design-o-meter: Towards Evaluating and Refining Graphic Designs

   Sahil Goyal, Abhinav Mahajan, Swasti Mishra, and 4 more authors

   In 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Jun 2025

   Abs DOI

   Graphic designs are an effective medium for visual communication. They range from greeting cards to corporate flyers and beyond. Off-late, machine learning techniques are able to generate such designs, which accelerates the rate of content production. An automated way of evaluating their quality becomes critical. Towards this end, we introduce Design-o-meter, a data-driven methodology to quantify the goodness of graphic designs. Further, our approach can suggest modifications to these designs to improve its visual appeal. To the best of our knowledge, Design-o-meter is the first approach that scores and refines designs in a unified framework despite the inherent subjectivity and ambiguity of the setting. Our exhaustive quantitative and qualitative analysis of our approach against baselines adapted for the task (including recent Multimodal LLM-based approaches) brings out the efficacy of our methodology. We hope our work will usher more interest in this important and pragmatic problem setting.

2023

1. paper

   AVA: AI-driven Virtual Rehabilitation Assistant

   Ali Abedi, Tracey JF Colella, Mark Bayley, and 6 more authors

   In 15th International Conference on Virtual Rehabilitation (WCISVR), Jun 2023

   Abs DOI

   Virtual rehabilitation has gained popularity in delivering personalized programs of exercise, education, and counseling to the home of patients. Despite the potential benefits of virtual rehabilitation programs in reducing rehospitalization and death, high dropout rates pose a significant obstacle to their effectiveness. This is due to several barriers, including a lack of motivation and confidence in completing rehabilitation exercises. This paper introduces an AI-driven Virtual Assistant (AVA) to assist patients in completing their prescribed rehabilitation exercises at home. AVA uses AI algorithms to analyze patients movements and provide them with real-time personalized feedback. The web application containing AVA can be accessed from any camera-enabled computer or mobile device without the need for additional hardware. Through a co-design approach, the movement training components of AVA for upper-limb stroke rehabilitation exercises were developed and reviewed by the research team, including a patient partner. The importance of including an avatar in virtual rehabilitation and providing realtime feedback to guide patients in performing exercises correctly was emphasized by the patient partner. AVA has the potential to enhance healthcare outreach, increase program participation and completion, and improve long-term health outcomes.