A5081210463- Topic Modeling
- Multimodal Machine Learning Applications
- Natural Language Processing Techniques
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Big Data and Business Intelligence
- Text and Document Classification Technologies
- Advanced Text Analysis Techniques
- Domain Adaptation and Few-Shot Learning
- Video Analysis and Summarization
- Semantic Web and Ontologies
- Web Data Mining and Analysis
- Privacy-Preserving Technologies in Data
- Advanced Database Systems and Queries
- Big Data Technologies and Applications
- Generative Adversarial Networks and Image Synthesis
- Machine Learning and Data Classification
- Sentiment Analysis and Opinion Mining
- Advanced Image Processing Techniques
- Advanced Graph Neural Networks
- Blockchain Technology Applications and Security
- Recommender Systems and Techniques
- Data Stream Mining Techniques
- Information Retrieval and Search Behavior
- Gene expression and cancer classification
Saarland University
2019-2021
Karlsruhe Institute of Technology
2015-2018
Iberia (Spain)
2017
International Institute of Information Technology, Hyderabad
2012-2013
Indian Institute of Technology Hyderabad
2011-2012
Interest in Artificial Intelligence (AI) and its applications has seen unprecedented growth the last few years. This success can be partly attributed to advancements made sub-fields of AI such as machine learning, computer vision, natural language processing. Much these fields been possible with deep a sub-area learning that uses artificial neural networks. created significant interest integration vision language. In this survey, we focus on ten prominent tasks integrate by discussing their...
In many languages, sparse availability of resources causes numerous challenges for textual analysis tasks.Text classification is one such standard tasks that hindered due to limited label information in lowresource languages.Transferring knowledge (i.e.label information) from high-resource low-resource languages might improve text as compared the other approaches like machine translation.We introduce BRAVE (Bilingual paRAgraph VEctors), a model learn bilingual distributed representations...
Human moods continuously change over time. Tracking can provide important information about psychological and health behavior of an individual. Also, history mood be used to predict the future individuals. In this paper, we try transition a Twitter user by regression analysis on tweets posted twitter time line. Initially, are automatically labeled with labels from 0 t-1. It is then at t. Experiments show that SVM attained less root-mean-square error compared other approaches for prediction.
Images in the wild encapsulate rich knowledge about varied abstract concepts and cannot be sufficiently described with models built only using image-caption pairs containing selected objects. We propose to handle such a task guidance of base that incorporate many concepts. Our method is two-step process where we first build multi-entity-label image recognition model predict as labels then leverage them second step an external semantic attention constrained inference caption generation for...
We wish to address the challenging task of opinion mining about organizations, people and places from different languages. It is known that resources tools for opinions are scarce. In our study, we leverage comparable news articles collection retrieve (opinion targets) in resource scarce language like Hindi. Opinions expressed targets (Named Entities)given by adjectives verbs as words extracted English corpora get transliterated translated scare Transformed then used create subjective model...
Social media platforms have grown into an important medium to spread information about event published by the traditional media, such as news articles. Grouping diverse sources of that discuss same topic in varied perspectives provide new insights. But gap word usage between informal social content tweets and diligently written (e.g. articles) make assembling difficult. In this paper, we propose a transformation framework bridge online articles across languages leveraging their embeddings....
Growth of multimodal content on the web and social media has generated abundant weakly aligned image-sentence pairs. However, it is hard to interpret them directly due intrinsic intension. In this paper, we aim annotate such pairs with connotations as labels capture We achieve a connotation embedding model (CMEM) using novel loss function. It's unique characteristics over previous models include: (i) exploitation data opposed only visual information, (ii) robustness outlier in multi-label...
Given an image, generating its natural language description (i.e., caption) is a well studied problem. Approaches proposed to address this problem usually rely on image features that are difficult interpret. Particularly, these subdivided into global and local features, where extracted from the representation of while objects detected locally in image. Although, extract rich visual information existing models generate captions blackbox manner humans have difficulty interpreting which caption...