Chirag Bhuvaneshwara

Experience

• 

  Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI)

  Aug 2019 - Jul 2024

  I work with the Affective Computing Group to facilitate emotion modeling for more realistic and involved user interaction with virtual characters. To this end, I help in further improving the Visual Scenemaker software to enable psychologists to easily model the interactions. Other than this, I work on applying machine learning to Social Signal Processing tasks such as face expression recognition. I have set up several interactive agent systems with multiple project partners and conducted studies, several of which have also recorded audio visual data for machine learning purposes. I am also working on annotation schemes for the collected data. Show less At the Affective Computing Group, I worked on the Visual Scenemaker to add plugins that allows the Java backend to communicate with an Android App where a user interacts with an avatar. I also developed responsive, mobile GUIs to direct conversations between the users and the virtual agent. The virtual agent was aimed at reliving stress in the EMMA project.Additionally, I decided on a Software package for setting up a Dialogue System for helping a person work with large robot arms used in assembly lines in the Mindbot project. In addition to the GUI, I will be working on the implementation here. Show less Working with the Interactive Machine Learning Group. Interactive Object Annotation Tool: Created a browser based tool that allows a user to upload images and have them pre-annotated using Facebook’s detectron based Mask RCNN model and then the GUI allows for the editing of the annotations as needed by the user. The GUI for the app was developed in two stages. The final GUI (v2) was developed with ReactJSI initially worked on classifying skin lesions images and some accompanying metadata into skin cancer classes using deep learning. My contribution has been towards changing the model architecture and making use of the metadata to combine neural network outputs and classical Machine Learning models. Show less

  • Researcher

    Apr 2022 - Jul 2024

  • Junior Researcher

    Aug 2021 - Apr 2022

  • Full Stack Developement Research Assistant

    Jul 2020 - Aug 2021

  • Machine Learning Research Assistant

    Aug 2019 - Aug 2020
• 

  The Affective Computing Group

  Apr 2022 - Jul 2024

  Researcher
• 

  K8 Institut für strategische Ästhetik

  Aug 2024 - now

  Data Scientist

  I work with K8 in the Twin4Resilience EU project to analyse Digital Twin City systems. We are preparing a comprehensive training program to educate stakeholders to better teach various sub-teams in the city management spehere about the concepts of Digital Twins. My role particularly is to develop the technical design modules of the training programme which will educate various stakeholders on the many key concepts invlovled in Digital Twin systems.