arXiv subjects

Garvita Allabadi

Publications and source records attributed to Garvita Allabadi.

3 recordsLinked to original sources

arXiv2026

GRIP: Feedback-Guided Prompt Retrieval for Large Multimodal Models

In-Context Learning (ICL) has become a powerful mechanism for adapting Large Language Models (LLMs) to new tasks without fine-tuning. Extending this concept to Large Multimodal Models (LMMs), Multimodal In-Context Learning (M-ICL) relies on retrieving relevant examples, such as images, captions, or question-answer pairs, to guide predictions across tasks like classification, captioning, and visual question answering (VQA). Most existing approaches select in-context examples based on feature-space similarity, assuming that semantically similar samples provide the most useful context. However, our systematic analysis reveals that this assumption does not always hold: visually similar examples are not necessarily those that most effectively enhance in-context learning performance. To address this, we propose the Guided Retrieval of In-context Prompts (GRIP), a learnable vision-only retrieval framework that leverages feedback from LMMs to identify examples that truly improve model predictions. GRIP learns to distinguish beneficial from detrimental in-context examples through contrastive training, refining retrieval beyond pure similarity. Across three multimodal tasks, namely classification, captioning, and VQA, GRIP improves consistently over similarity-based retrieval on Qwen2.5-VL-7B, with its strongest gains in classification on Idefics2-8B. Moreover, we demonstrate that retrievers trained with feedback from one open LMM can be transferred to other models without retraining, including closed-source GPT-4o and Gemini, enabling scalable and cost-efficient deployment of M-ICL. Code will be published upon acceptance.

cs.CV↗

arXiv2024

Generalized Open-World Semi-Supervised Object Detection

Traditional semi-supervised object detection methods assume a fixed set of object classes (in-distribution or ID classes) during training and deployment, which limits performance in real-world scenarios where unseen classes (out-of-distribution or OOD classes) may appear. In such cases, OOD data is often misclassified as ID, thus harming the ID classes accuracy. Open-set methods address this limitation by filtering OOD data to improve ID performance, thereby limiting the learning process to ID classes. We extend this to a more natural open-world setting, where the OOD classes are not only detected but also incorporated into the learning process. Specifically, we explore two key questions: 1) how to accurately detect OOD samples, and, most importantly, 2) how to effectively learn from the OOD samples in a semi-supervised object detection pipeline without compromising ID accuracy. To address this, we introduce an ensemble-based OOD Explorer for detection and classification, and an adaptable semi-supervised object detection framework that integrates both ID and OOD data. Through extensive evaluation on different open-world scenarios, we demonstrate that our method performs competitively against state-of-the-art OOD detection algorithms and also significantly boosts the semi-supervised learning performance for both ID and OOD classes.

cs.CV↗

arXiv2021

smol: Sensing Soil Moisture using LoRa

Technologies for environmental and agricultural monitoring are on the rise, however, there is a lack of small, low-power, and lowcost sensing devices in the industry. One of these monitoring tools is a soil moisture sensor. Soil moisture has significant effects on crop health and yield, but commercial monitors are very expensive, require manual use, or constant attention. This calls for a simple and low-cost solution based on novel technology. In this work, we introduce smol: Sensing Soil Moisture using LoRa, a low-cost system to measure soil moisture using received signal strength indicator (RSSI) and transmission power. It is compact and can be deployed in the field to collect data automatically with little manual intervention. Our design is enabled by the phenomenon that soil moisture attenuates wireless signals, so the signal strength between a transmitter-receiver pair decreases. We exploit this physical property to determine the variation in soil moisture. We designed and tested our measurement-based prototype in both indoor and outdoor environments. With proper regression calibration, we show soil moisture can be predicted using LoRa parameters.

eess.SP↗