Faculty Spotlight: Heng Ji
Heng Ji is a professor at Computer Science Department and part of Thrust 1 in MMLI. Her esearch focuses on Natural Language Processing, especially on Multimedia Multilingual Information Extraction, Knowledge-enhanced Large Language Models and Vision-Language Models, AI for Science and Science-inspired AI. There exist approximately 166 billion small molecules, with 970 million deemed druglike. Despite this vast pool, only a tiny portion of molecules are currently approved across global healthcare systems. This scarcity underscores the urgent need for innovative approaches, motivating me to contribute significantly to molecule function discovery with advanced AI technologies. Many of these technologies can be applied to other domains such as new material discovery. [Click above to read more!]
Trainee Spotlight: Bryan Amador
Bryan Amador Bryan is a graduate student in the Computing and Information Science PhD program at Rochester Institute of Technology, where he is currently in his third year and a member of the Document and Pattern Recognition Lab (DPRL). In the DPRL lab, they research systems that recognize and retrieve information from documents, images, and videos. Bryan’s area of research is Math Information Retrieval, where the goal is to develop tools that facilitate users’ access to mathematical content. Right now, he is most excited to work on modeling the task of formula search as a graph embedding and retrieval problem taking advantage of the visual and semantic content of formulas. [Click above to read more!]
MMLI Staff Spotlight: Heng Ji
What is your background and describe your current work/role/any exciting projects you want to share.
My research focuses on Natural Language Processing, especially on Multimedia Multilingual Information Extraction, Knowledge-enhanced Large Language Models and Vision-Language Models, AI for Science and Science-inspired AI.
What drew you to MMLI? What has been your favorite part of being a part of MMLI?
It’s the people. It’s a wonderful privilege to be able to work with wonderful chemists such as Profs. Martin Burke, Ying Diao and Huimin Zhao. They are extremely intelligent, open-minded, kind and generous, and have become my close collaborators and good friends.
How do you like to spend your free time? I regularly do yoga and Zumba indoor, and have been enjoying outdoor activities in the countryside, especially cycling and hiking in the forests. I have been traveling to 30+ countries in six continents.
Fun fact (or extremely average fact) about yourself. I’m very creative in terms of making dumplings.
MMLI Trainee Spotlight: Let’s meet Bryan Amador
What is your background and describe your current work/role/lab and the project you are most excited to be working on right now.
I was born in Tegucigalpa, Honduras. In 2021, I got a bachelor in Computing Systems Engineering at the Universidad Tecnológica Centroamericana (UNITEC). After graduating, I worked as a backend developer at Grupo Vesta and as a frontend developer at Soluciones Digitales en Salud. In Fall 2022, I moved to the United States to start in the Computing and Information Science PhD program at Rochester Institute of Technology, where I am currently in my third year. I joined as a member of the Document and Pattern Recognition Lab (DPRL). In our lab, we research systems that recognize and retrieve information from documents, images, and videos. My area of research is Math Information Retrieval, where the goal is to develop tools that facilitate users’ access to mathematical content. At the same time, I collaborate with projects of information recognition being held in the lab. Right now, I am most excited to work on modeling the task of formula search as a graph embedding and retrieval problem taking advantage of the visual and semantic content of formulas.
What drew you to your project and/or MMLI?
Our lab has been working by recognizing math formulas from documents. That is, given a formula image, infer the symbols and the structure of them so that they can be used in downstream tasks as retrieval. So we thought similar techniques can be applied to molecular diagrams. With this in mind, ChemScraper was created. The main goal of this tool is to extract molecules from documents and export them to formats that can be used by chemists for their needs. For instance, produce SMILES from molecules in chemical papers that can be further indexed and searched to facilitate reaction planning.
What has been your favorite part of being a part of MMLI?
The in-person meetings (councils, retreats, etc) are my favorite part of MMLI, they bring together people from different positions in all entities affiliated to the institute (students, professors, industry managers, etc). Being such a big institute, I was able to meet people from different backgrounds working on interesting problems. I got useful advice about short and long term careers in both industry and academia from successful people in both worlds. Some activities allowed me to not just hear from these people from a podium (which is fruitful as well), but to sit face to face and share experiences and ideas, giving me the opportunity to meet the actual person that is behind an important position.
How do you like to spend your free time? (or what would you do for fun if you had more free time!)
I love to play football, either with people or practicing by myself. If I had more time, I would like to continue learning to play guitar (my old guitar is probably full of dust back home). I like to visit places where I can appreciate natural wonders. I like to cook and try new dishes whenever I can. I sometimes play video games as well.
Fun fact (or extremely average fact) about yourself you would like to share.
I love tulips. I wish one day I can have my own tulip garden.