V
V. Vinay
Publications - 8
Citations - 33
V. Vinay is an academic researcher. The author has contributed to research in topics: Computer science & Chemistry. The author has an hindex of 1, co-authored 1 publications receiving 2 citations.
Papers
More filters
Proceedings ArticleDOI
Curriculum Learning for Dense Retrieval Distillation
Hansi Zeng,Hamed Zamani,V. Vinay +2 more
TL;DR: A generic curriculum learning based optimization framework called CL-DRD that controls the difficulty level of training data produced by the re-ranking (teacher) model is proposed that iteratively optimizes the dense retrieval (student) model by increasing the difficulty of the knowledge distillation data made available to it.
Proceedings Article
VarScene: A Deep Generative Model for Realistic Scene Graph Synthesis
TL;DR: In this article , a variational autoencoder for scene graphs is proposed, which is optimized for the maximum mean discrepancy (MMD) between the ground truth scene graph distribution and distribution of generated scene graphs.
Proceedings ArticleDOI
Robustness of Fusion-based Multimodal Classifiers to Cross-Modal Content Dilutions
Gaurav Verma,V. Vinay +1 more
TL;DR: In this article , the authors investigate the robustness of multimodal classifiers to cross-modal dilutions and develop a model that generates additional dilution text that maintains relevance and topical coherence with the image and existing text.
Dissertation
The relevance of feedback for text retrieval
TL;DR: An evaluation framework for measuring the effectiveness of feedback algorithms and properties that characterise sets of documents are proposed with the particular aim of identifying measures that are predictive of future performance of statis tical algorithms on these document sets.
Proceedings ArticleDOI
NeurTEx: A Neural Framework for Template Extraction from Flat Images
Vinay Aggarwal,Praneetha Vaddamanu,Bhanu Prakash Reddy Guda,Balaji Vasan Srinivasan,Niyati Chhaya,V. Vinay +5 more
TL;DR: This work proposes NeurTEx, a holistic algorithm that takes an inspirational banner image as input and extracts multimodal design semantics: layout, text elements, image elements, logos and background/foreground objects and shapes, and demonstrates how these extractions can accelerate the creation process thus aiding novices and professionals alike.