Abstract

Scene understanding and reasoning has been a fundamental problem in 3D computer vision, requiring models to identify objects, their properties, and spatial or comparative relationships among the objects. Existing approaches enable this by creating scene graphs using multiple inputs such as 2D images, depth maps, object labels, and annotated relationships from specific reference view. However, these methods often struggle with generalization and produce inaccurate spatial relationships like "left/right", which become inconsistent across different viewpoints. To address these limitations, we propose Viewpoint-Invariant ZerO-shot scene graph generation for 3D scene Reasoning (VIZOR). VIZOR is a training-free, end-to-end framework that constructs dense, viewpoint-invariant 3D scene graphs directly from raw 3D scenes. The generated scene graph is unambiguous, as spatial relationships are defined relative to each object’s front-facing direction, making them consistent regardless of the reference view. Furthermore, it infers open-vocabulary relationships that describe spatial and proximity relationships among scene objects without requiring annotated training data. We conduct extensive quantitative and qualitative evaluations to assess the effectiveness of VIZOR on scene graph generation and downstream tasks, such as query-based object grounding. VIZOR outperforms state-of-the-art methods, showing clear improvements in scene graph generation and achieving 22% and 4.81% gains in zero-shot grounding accuracy on the Replica and Nr3D datasets, respectively.

Abstract

Generating realistic human motions that naturally respond to both spoken language and physical objects is crucial for interactive digital experiences. Current methods, however, address speech-driven gestures or object interactions independently, limiting real-world applicability due to a lack of integrated, comprehensive datasets. To overcome this, we introduce InteracTalker, a novel framework that seamlessly integrates prompt-based object-aware interactions with co-speech gesture generation. We achieve this by employing a multi-stage training process to learn a unified motion, speech, and prompt embedding space. To support this, we curate a rich human-object interaction dataset, formed by augmenting an existing text-to-motion dataset with detailed object interaction annotations. Our framework utilizes a Generalized Motion Adaptation Module that enables independent training, adapting to the corresponding motion condition, which is then dynamically combined during inference. To address the imbalance between heterogeneous conditioning signals, we propose an adaptive fusion strategy, which dynamically reweights the conditioning signals during diffusion sampling. InteracTalker successfully unifies these previously separate tasks, outperforming prior methods in both co-speech gesture generation and object-interaction synthesis, outperforming gesture-focused diffusion methods, yielding highly realistic, object-aware full-body motions with enhanced realism, flexibility, and control.(https://sreeharirajan.github.io/projects/InteracTalker/)

Abstract

Early-stage fruit yield prediction plays a key role in supporting timely agronomic decisions, enhancing market planning, and empowering farmers with data-driven insights. Over the years, most approaches to yield estimation have focused on fruit counting techniques, typically performed just before harvest. While these methods have proven useful, they often come into play late in the cultivation cycle, limiting their impact on early planning and resource optimization. In this work, we introduce a comprehensive baseline framework for predicting mango yield at an earlier stage - during flowering - using image-based learning. Our contributions are twofold. (i) Our approach combines a SegFormer-based segmentation model with a regression pipeline to estimate yield from images, while also exploring the role of contextual features such as weather and scale. (ii) This work introduces a novel benchmark and an enriched dataset, paving the way for scalable, automated tools that can assist farmers and stakeholders in making proactive decisions throughout the mango growing season. Our work demonstrates that for multi-modal yield prediction, integrating features that complement visual representations (like scale) can be more impactful than using features with a stronger standalone linear correlation (like weather). Our single-image model, based on the SegFormer-B1 encoder, achieved a mean absolute error (MAE) of 7.68, R² of 0.76, and mean squared error (MSE) of 115.48. These results highlight the promise of vision-based models for yield estimation from early-stage flowering cues. To the best of our knowledge, this is the first work to address the prediction of mango yield using images from the flowering stage and weather data.

Abstract

Temporal Graph Neural Networks (TGNNs) are increasingly used in high-stakes domains, such as financial forecasting, recommendation systems, and fraud detection. However, their susceptibility to poisoning attacks poses a critical security risk. We introduce LORETTA (Low Resource Twophase Temporal Attack), a novel adversarial framework on Continuous-Time Dynamic Graphs, which degrades TGNN performance by an average of 29.47% across 4 widely benchmark datasets and 4 State-of-the-Art (SotA) models. LORETTA operates through a two-stage approach: (1) sparsify the graph by removing high-impact edges using any of the 16 tested temporal importance metrics, (2) strategically replace removed edges with adversarial negatives via LORETTA’s novel degree-preserving negative sampling algorithm. Our plug-and-play design eliminates the need for expensive surrogate models while adhering to realistic unnoticeability constraints. LORETTA degrades performance by upto 42.0% on MOOC, 31.5% on Wikipedia, 28.8% on UCI, and 15.6%on Enron. LORETTA outperforms 11 attack baselines, remains undetectable to 4 leading anomaly detection systems, and is robust to 4 SotA adversarial defense training methods, establishing its effectiveness, unnoticeability, and robustness.

Abstract

A system and a method for recommending background music that corresponds to an extracted text from a book based on emotion and a topic that is relevant to the extracted text using machine learning models provided. The method includes,(i) determining, using a first trained machine learning model, the emotion from the extracted text that corresponds to the paragraph of the book,(ii) assigning, using a word similarity technique, a similarity score for

Abstract

Personalized Interactive Music Systems (PIMSs) are emerging as promising devices for enhancing physical activity and exercise outcomes. By leveraging real-time data and adaptive technologies, PIMSs align musical features, such as tempo and genre, with users’ physical activity patterns, including frequency and intensity, enhancing their overall experience.

Abstract

Background Prior research by the authors studied the objective impact on medical students’ academic course, the perceived cognitive load, quantified subjective feedback while teaching using AnaVu, a low resource stereoscopic projection system. The qualitative data from opinions, and comparative educational results reported by undergraduate medical students following their participation in stereoscopic (AnaVu) and monoscopic learning sessions are explored in this study. This study was done to evaluate and compare the effectiveness of stereoscopic and monoscopic teaching methods in enhancing spatial understanding of anatomical structures among first-year undergraduate medical students using AnaVu. Methodology This research study was conducted as a three-limb randomized controlled trial. Among those who provided informed consent, a sample of MBBS students from the 2022 cohort was chosen at random. Following a one-hour brainstem introduction lecture and a dissection session, students were assigned at random to one of three groups: S for stereo; M for mono; or C for control. A 20-minute stereoscopic demonstration of the brainstem module in AnaVu has been delivered to S. The identical presentation, albeit in monoscopic mode, was delivered to M. Diagrams drawn on a whiteboard were used to instruct the C group. Pre-intervention and post-intervention tests were given in four domains: basic recall, analytical, radiological anatomy and diagram-based questions. Finally, the groups were swapped – S→M, M→S and C→S, and they were asked to compare the modes. Data were processed using RQDA (R package for Qualitative Data Analysis). Theme identification and analysis of the qualitative data was done using the thematic analysis. Results 152 students participated in the study. Five themes and fourteen subthemes were identified. General advantages included size comparisons, software features, and improved attention, while specific benefits for learning radiological and sectional anatomy were noted. Disadvantages included eye strain, software glitches, costs, and concerns about teacher adaptation. Conclusion The findings of this study emphasize the potential of AnaVu in enhancing anatomical and radiological education, while highlighting key disadvantages to help teachers and students make informed choices between stereo and mono display methods.

Abstract

Display advertising is one of the predominant modes of online advertising. Given a set of available ad slots on the web pages, a publisher makes efforts to allocate them to advertisers, satisfying their demand and maximizing their revenue. Investigating efficient approaches for ad slot allocation to advertisers is a research issue. Given a set of ad slots from different pages, advertisers aim to deliver the ad to a specified number of targeted users. Exposing a user repeatedly to the same ad may result in a potential loss of revenue for the publisher and create boredom for the user. There is an opportunity to improve the publisher’s revenue by ensuring that each ad slot allocated to an advertiser reaches a distinct number of users. In the literature, efforts are being made to address this problem by modeling it as an optimization problem. These approaches suffer from the issues of complexity and practicality. We approach the problem with the pattern-mining framework by employing the notions of “coverage” and “overlap”. In the proposed approach, we exploit the knowledge of coverage patterns extracted from click stream transactions and propose a comprehensive, efficient, and practical ad slot allocation framework. The experimental results on two real-world click stream datasets show that the proposed approach could improve the revenue of the publisher by meeting the demands of the increased number of advertisers by reducing the repeated display of advertisements to users.

Abstract

n the realm of multi-armed bandit problems, the Gittins index policy is known to be optimal in maximizing the expected total discounted reward obtained from pulling the Markovian arms. In most realistic scenarios however, the Markovian state transition probabilities are unknown and therefore the Gittins indices cannot be computed. One can then resort to reinforcement learning (RL) algorithms that explore the state space to learn these indices while exploiting to maximize the reward collected. In this work, we propose tabular (QGI) and Deep RL (DGN) algorithms for learning the Gittins index that are based on the retirement formulation for the multi-armed bandit problem. When compared with existing RL algorithms that learn the Gittins index, our algorithms have a lower run time, require less storage space (small Q-table size in QGI and smaller replay buffer in DGN), and illustrate better empirical convergence to the Gittins index. This makes our algorithm well suited for problems with large state spaces and is a viable alternative to existing methods. As a key application, we demonstrate the use of our algorithms in minimizing the mean flowtime in a job scheduling problem when jobs are available in batches and have an unknown service time distribution.

Abstract

Dynamic pricing is the practice of adjusting the selling price of a product to maximize a firm’s revenue by responding to market demand. The literature typically distinguishes between two settings: infinite inventory, where the firm has unlimited stock and time to sell, and finite inventory, where both inventory and selling horizon are limited. In both cases, the central challenge lies in the fact that the demand function — how sales respond to price — is unknown and must be learned from data. Traditional approaches often assume a specific parametric form for the demand function, enabling the use of reinforcement learning (RL) to identify near-optimal pricing strategies. However, such assumptions may not hold in real-world scenarios, limiting the applicability of these methods. In this work, we propose a Gaussian Process (GP) based nonparametric approach to dynamic pricing that avoids restrictive modeling assumptions. We treat the demand function as a black-box function of the price and develop pricing algorithms based on Bayesian Optimization (BO)—a sample-efficient method for optimizing unknown functions. We present BO-based algorithms tailored for both infinite and finite inventory settings and provide regret guarantees for both regimes, thereby quantifying the learning efficiency of our methods. Through extensive experiments, we demonstrate that our BO-based methods outperform several state-of-the-art RL algorithms in terms of revenue, while requiring fewer assumptions and offering greater robustness. This highlights Bayesian Optimization as a powerful and practical tool for dynamic pricing in complex, uncertain environments.