CV

Education

• M.S. in Information Science, University of Pittsburgh, 2024
• BTech. in Computer Science and Engineering, Indian Institute of Information Technology, Nagpur, 2021

Research experience

• Stanford Intelligent Systems Laboratory
  • Summer 2023: Independent Student Researcher
  • Supervisor: Mansur Arief
    • Improving rare traffic sign recognition via data augmentation: Addressed challenges of out-of-training distribution images(eg., Rusty traffic signs), crucial for vehicle perception systems. Under review at ICRA.
    • Augmented the German Traffic Sign Recognition Benchmark dataset through neural style transfer, introducing rusty sign images. Constructed a pipeline with Variational Prototyping Encoders to identify and classify rare image class, achieving a 0.86 class-weighted F1 score.
• University of Pittsburgh
  • Summer 2023: Graduate Student Researcher
  • Supervisor: Daqing He
    • Guided summarization of clinical notes: Developed a comprehensive summary template covering patient demographics, chief complaint, OPQRST assessment, diagnostics, treatment, etc
    • Developed a sentence classifier to determine sentence importance within clinical notes, providing guidance to the BART-based summarization model.
    • Additionally, fine-tuned a Bio-ClinicalBERT-based Named Entity Recognition (NER) system, enabling the implementation of a fact-checking metric for validating predicted summaries against the ground truth.
    • Currently experimenting with medically focused prompt engineering techniques by running multi-GPU inference using Huggingface Accelerate on large language models (LLMs) like LLaMA to improve the task of clinical note summarization.
• University of California, San Diego
  • 2021-2022: Research Intern
  • Advisor: Pengtao Xie
    • Neural Architecture Search (NAS) for Pneumonia Diagnosis: Implemented Neural Architecture Search (NAS) for improving pneumonia diagnosis from Chest X-Ray images by Leveraging the ’Learning By Teaching’ framework, inspired by teacher-student learning paradigm that outperforms previous NAS methods like DARTS and PC-DARTS by 5.1%
    • The searched model attained a 97.6% ROC-AUC score for pneumonia detection, while being 4% smaller than DARTS.
    • Reading by Translating: Implemented the ’Reading by Translating’ framework that improves the task of ’Machine Reading’ i.e extracting meaningful instances from the dataset.
    • Involves 2 transformer based encoder-decoder models, that are trained mutually on the task of Machine Translation to learn importance weights assigned to the dataset instances.
• Tata Research, Design and Development Centre
  • 2020: Research and Development Intern
  • Advisor: Sagar Sunkle
    • Named Entity Recognition (NER): Worked on information retrieval using NER on domain-specific scientific corpora.
    • Implemented incremental learning for NER by adapting classifiers from open-source libraries such as Adaptive Random Forests, AdaBoost, Pretrained Spacy NER, Conditonal Random Fields (CRF) and Seq2Seq model.
    • Developed an ’Extractive search system for entity retrieval’ similar to AllenAI’s Spike cord search system that enables one to perform selective information extraction and annotate results by querying
• National Institute of Technology, Nagpur
  • 2018: Research Intern
  • Advisor: Dr. Mansi A Radke
    • Word sense disambiguation: Detected the ‘geospatialness’ of a preposition in a sentence.
    • Formulated a geo-parser using NLP (implemented in Python) preprocessing tools that performed gazetteer lookups to extract placenames and geo-feature types from a sentence.
    • Implemented baseline classification methods like Naive Bayes, BayesNet, Random Forests, SVM’s and meta-classifiers.
    • Conducted a comparative study by fine-tuning pre-trained models like ULMFiT, XLNet, Transformer-XL, RoBERTa, BERT, ALBERT and DistilBERT.
    • Achieved an F1 score of 0.947, an increase of 15% as compared to baseline methods

Skills

• Languages - Python, R, MySQL, C++, C, Bash
• ML/DL - PyTorch, Tensorflow, Fastai, Keras, scikit-learn, spaCy, Hugging Face, Langchain, Caret, glmnet
• Data Processing - Pandas, NLTK, CoreNLP, Gensim, Tidyverse (Dplyr, Tibble, Purrr), igraph
• Visualisation - Matplotlib, Seaborn, ggplot
• Graduate Coursework: Machine Learning in R (Code), Databases, Algorithms, Data Mining, Artificial Intelligence, Information Storage and Retrieval
• Undergraduate Coursework: Natural Language Processing, Deep Learning, Computer Vision, Neuro-Fuzzy Techniques, Probability, Graph Theory, Data Science Bio-informatics

Publications

Neural Architecture Search for Pneumonia Diagnosis from Chest X-rays Permalink

Gupta, A., Sheth, P. & Xie, P. Neural architecture search for pneumonia diagnosis from chest X-rays. Sci Rep 12, 11309 (2022). https://doi.org/10.1038/s41598-022-15341-0.

Disambiguating Spatial Prepositions: The Case of Geo-spatial Sense Detection Permalink

Radke, M. A., Gupta, A., Stock, K., & Jones, C. B. (2022). Disambiguating spatial prepositions: The case of geo-spatial sense detection. Transactions in GIS, 00, 1–31. https://doi.org/10.1111/tgis.12976

Towards Accurate and Clinically Meaningful Summarization of Electronic Health Record Notes: A Guided Approach Permalink

@inproceedings{luo2023towards, title={Towards Accurate and Clinically Meaningful Summarization of Electronic Health Record Notes: A Guided Approach}, author={Luo, Zhimeng and Ji, Yuelyu and Gupta, Abhibha and Li, Zhuochun and Frisch, Adam and He, Daqing}, booktitle={2023 IEEE EMBS International Conference on Biomedical and Health Informatics (BHI)}, pages={1--5}, year={2023}, organization={IEEE} }

Enhancing Visual Perception in Novel Environments via Incremental Data Augmentation Based on Style Transfer Permalink

@misc{gupta2023enhancing, title={Enhancing Visual Perception in Novel Environments via Incremental Data Augmentation Based on Style Transfer}, author={Abhibha Gupta and Rully Agus Hendrawan and Mansur Arief}, year={2023}, eprint={2309.08851}, archivePrefix={arXiv}, primaryClass={cs.CV} }

Argumentative Stance Prediction: An Exploratory Study on Multimodality and Few-Shot Learning Permalink

@inproceedings{sharma-etal-2023-argumentative, title = "Argumentative Stance Prediction: An Exploratory Study on Multimodality and Few-Shot Learning", author = "Sharma, Arushi and Gupta, Abhibha and Bilalpur, Maneesh", editor = "Alshomary, Milad and Chen, Chung-Chi and Muresan, Smaranda and Park, Joonsuk and Romberg, Julia", booktitle = "Proceedings of the 10th Workshop on Argument Mining", month = dec, year = "2023", address = "Singapore", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.argmining-1.18", doi = "10.18653/v1/2023.argmining-1.18", pages = "167--174" }

Talks

• Conducted a day long workshop on ‘Introduction to Machine Learning’ at the Cummins college of Engineering, Nagpur.