What am I doing now?

I am currently working as a researcher at Smart Information Flow Technologies (SIFT). In my spare time, I work on side projects at the intersection of langauge, vision, and representation learning to better understand these areas. This interest was motivated by a recent project I worked on with image captioning models. If you have any suggested papers to look at, feel free to contact me.

Who am I?

Hello! I'm a researcher at Smart Information Flow Technologies (SIFT), working at the intersection of machine learning, deep learning, NLP, and AI planning and plan recognition. I also do photography in my spare time, so feel free to check out my photo gallery once it's set up!

If you're interested in contacting me, I can be reached at pavank92 at gmail dot com

Education

  • PhD - Computer Science, Sept. 2014 - June 2021
    Department of Computer Science, Drexel University, Philadelphia, PA
    Supervisor: Dr. Santiago Ontañón
    Thesis: Learning Decomposition Models for Hierarchical Planning and Plan Recognition
  • M.S. - Computer Science, Sept. 2014 - June 2017
    Department of Computer Science, Drexel University, Philadelphia, PA
    Supervisor: Dr. Christopher Geib
  • B.S. - Computer Science, Sept. 2010 - June 2014
    Department of Computer Science, Drexel University, Philadelphia, PA
    Specialization: Artificial Intelligence and Computing Systems
    Minor: Computer Engineering

Research

My research interests span several subfields including machine/deep learning, NLP, and planning. I have 7+ years of experience working in symbolic AI planning and plan recognition, and symbolic machine learning. More recently, I have begun transitioning into the areas of Natural Language Processing and deep learning research.

Research Areas:

  • Machine Learning
  • Deep Learning
  • Natural Language Processing
  • AI Planning
  • AI Planning Recognition

Publications

Blog Post

  1. Kantharaju, P., & Sankar, A. (2022). An Understanding of Learning from Demonstrations for Neural Text Generation. ICLR Blog Track. https://iclr-blog-track.github.io/2022/03/25/text-gen-via-lfd/
  2. Kantharaju, P. (2023). Captioning Pokémon Cards with Image-to-Text Models. Weights and Biases Fully Connected. https://api.wandb.ai/links/pkthunder/93zvdw8r

Journal Articles

  1. (Code) Rabkina, I., Kantharaju, P., Wilson, J. R., Roberts, M., & Hiatt, L. M. (2022). Evaluation of Goal Recognition Systems on Unreliable Data and Uninspectable Agents. Frontiers in Artificial Intelligence, 4. https://doi.org/10.3389/frai.2021.734521.
  2. Kantharaju, P., Alderfer, K., Zhu, J., Char, B., Smith, B., & Ontañón, S. (2022). Modeling Player Knowledge in a Parallel Programming Educational Game. IEEE Transactions on Games, 14(1), 64–75. https://doi.org/10.1109/TG.2020.3037505
  3. Manousakis, K., Eswaran, S., Shur, D., Naik, G., Kantharaju, P., Regli, W., & Adamson, B. (2015). Torrent-based Dissemination in Infrastructure-less Wireless Networks. Journal of Cyber Security and Mobility, 4(1), 1–22.

Conference Proceedings

  1. Rabkina, I., Kantharaju, P., Roberts, M., Wilson, J., Forbus, K., & Hiatt, L. M. (2020). Recognizing the Goals of Uninspectable Agents. Advances in Cognitive Systems.
  2. Kantharaju, P., & Ontañón, S. (2020). Discovering Meaningful Labelings for RTS Game Replays via Replay Embeddings. 2020 IEEE Conference on Games (CoG), 160–167.
  3. Kantharaju, P., Ontañón, S., & Geib, C. W. (2019). Scaling up CCG-Based Plan Recognition via Monte-Carlo Tree Search. 2019 IEEE Conference on Games (CoG), 1–8.
  4. Kantharaju, P., Alderfer, K., Zhu, J., Char, B., Smith, B., & Ontanón, S. (2018). Tracing Player Knowledge in a Parallel Programming Educational Game. Fourteenth Artificial Intelligence and Interactive Digital Entertainment Conference (AIIDE), 173–179.
  5. Kantharaju, P., Ontañón, S., & Geib, C. W. (2018). microCCG, a CCG-based Game-Playing Agent for microRTS. 2018 IEEE Conference on Computational Intelligence and Games, 1–8.
  6. Geib, C. W., & Kantharaju, P. (2018). Learning Combinatory Categorial Grammars for Plan Recognition. Thirty-Second AAAI Conference on Artificial Intelligence, 3007–3014.
  7. Geib, C., Weerasinghe, J., Matskevich, S., Kantharaju, P., Craenen, B., & Petrick, R. P. A. (2016). Building Helpful Virtual Agents Using Plan Recognition and Planning. Twelfth Artificial Intelligence and Interactive Digital Entertainment Conference (AIIDE), 162–168.

Workshop Papers

  1. Kantharaju, P., & Schmer-Galunder, S. (2022). Extracting Associations of Intersectional Identities with Discourse about Institution from Nigeria. NLPCSS.
  2. Rabkina, I., Kantharaju, P., Wilson, J., Roberts, M., & Hiatt, L. (2021). Comparing Hierarchical Goal Recognition via HTN, CCG, and Analogy. Workshop on Plan, Activity, and Intent Recognition (PAIR).
  3. Kantharaju, P., Ontañón, S., Geib, C., & Roberts, M. (2020). Combinatory Categorial Grammar Learning for Plan Recognition in Domains with Type Trees. Workshop on Plan, Activity, and Intent Recognition (PAIR) 2020.
  4. Rabkina, I., Kantharaju, P., Roberts, M., Wilson, J., Forbus, K., & Hiatt, L. (2020). Recognizing the Goals of Uninspectable Agents. Workshop on Plan, Activity, and Intent Recognition (PAIR) 2020.
  5. Kantharaju, P., Ontañón, S., & Geib, C. W. (2019). Extracting CCGs for Plan Recognition in RTS games. Workshop on Knowledge Extraction in Games 2019.

Research Code

  1. µCCG - Game-Playing Agent used in CIG 2018 µRTS Competition

Personal Projects

In my spare time, I like to work on side projects. I'll make these public when I get the chance!

Discord Chat Bot

I developed a Discord Chat Bot for a discord I frequent. This was one of my first serious side projects that I took on.

Pokémon Card Image Captioner

For the Effective MLOps - Model Development course from Weights and Biases, I trained and evaluated an image-to-text generation model for captioning Pokémon Cards.