News

2022

  • (Oct) Richard Zanibbi gave a talk for the Topos Institute on Mathematical Information Retrieval on Oct. 27th. The Topos colloquium series, including a YouTube link for the talk may be found online here: https://topos.site/topos-colloquium.
  • (Aug) Behrooz Mansouri's paper on contextualized formula search using MathAMR has been accepted for publication at CIKM 2022.
  • (Aug) Behrooz Mansouri defended his dissertation on math-aware search on July 26th, and submitted the final document in early August. Congratulations to Behrooz on completing his PhD! The dissertation is available here.
  • (July) ARQMath-3 is now complete. A big thanks to all who participated, and to our student assessors from RIT.
  • (May) As part of an independent study in Spring 2022, JP Ramissini has created a new YOLO-based detector for math formulas and chemical diagrams (GitLab link).
  • (May) Abhisek Dey is completing an internship at the University of Illinois this summer (as part of the MMLI project).
  • (May) Ayush Kumar Shah will be completing an internship with Amazon this summer as an Applied Research Scientist intern in Computer Vision / Deep Learning.
  • (May) Shaurya Rohatgi is completing another internship at AllenAI this summer.
  • (May) Congratulations to Matt Langsenkamp, who has been admitted to the PhD program at RIT.
  • (May) Congratulations to Abhisek Dey, who succesfully completed his comprehensive (Research Potential Assessment) for the RIT PhD program.
  • (March) Congratulations to Behrooz Mansouri, who will be holding a tenure-track faculty position in the Computer Science Department at the University of Southern Maine!

2021

  • (Dec) The Alfred P. Sloan foundation has funded a small grant in support of completing the project.
  • (Nov) Behrooz Mansouri has successfully defended his PhD dissertation proposal on improving use of surrounding context in formula search, and related work in math question answer retrieval.
  • (Oct) A cleaner and faster version of Symbol Scraper is now available: https://gitlab.com/dprl/symbolscraper-server
  • (Sept) The first version of the MathSeer extraction pipeline, which locates and recognizes formulas in PDF documents is now available through GitLab, just in time for Ayush K. Shah and Abhisek Dey's talk and poster presentations at GREC 2021 and ICDAR 2021.
  • (June - July ) The ARQMath-2 math-aware search task at CLEF 2021 is running! Our thanks to Behrooz Mansouri who has led the assessment effort, and managed the data collection and associated tools.
  • (May) MathDeck was demonstrated in an interactivity session at CHI 2021. Congratulations to Yancarlos Diaz, Gavin Nishizawa, and Behrooz Mansouri, the student authors on the paper.
  • (May) Yancarlos Diaz and Robin Avenoso have successfully defended their MSc theses! Their thesis documents are available here.
  • (May) Shaurya Rohatgi is completing an internship at Allen AI this summer.
  • (April) Behrooz Mansouri's paper on the first learning-to-rank model for formula search has been accepted for publication at SIGIR 2021, and will be participating in the SIGIR 2021 Doctoral Consortium.
  • (April) Ayush Kumar Shah and Abhisek Dey's paper on the MathSeer formula extraction and evaluation pipeline has been accepted for publication at ICDAR 2021.
  • (April) Prof. Zanibbi gave a seminar talk on the MathSeer project in the Boise State Computing PhD seminar series.
  • (March) ARQMath was presented at ECIR 2021 by Behrooz Mansouri (video).
  • (February) MathDeck will be demonstrated in an interactivity session at CHI 2021. Congratulations to Yancarlos Diaz, Gavin Nishizawa, and Behrooz Mansouri, the student authors on the paper.

2020

  • (December) We welcome Matt Langsenkamp to the project, who will be working as a research programmer in the dprl at RIT.
  • (December) We welcome Abhisek Day to the project. Abhisek is a new PhD student who recently completed his Master's in Computer Engineering at RIT.
  • (September) Yancarlos Diaz (BSc/MSc student) has received a job offer from Facebook. Yancarlos plans to work in their NYC office after graduation.
  • (August) Congratulations to Mahshad Mahdavi, who successfully defended her dissertation in Imaging Science (title: Query-Driven Global Graph Attention Model for Visual Parsing: Recognizing Handwritten and Typeset Math Formulas)
  • (August) Congratulations to Gavin Nishizawa, who successfully defended his Master's thesis in Computer Science (title: Visual Structure Editing of Math Formulas)
  • (August) We welcome Ayush Shah to the project. Ayush is a new RIT PhD student who will work on formula recognition and retrieval.
  • (June) RIT has released a story about MathDeck and the MathSeer project (link). The story was eventually carried by ACM TechNews, Phys.org, and other online blobs and news sources.
  • (May) We welcome Alex Keller to the project. Alex is a new RIT PhD student who will be working on formula indexing and retrieval.
  • (Apr) The MathDeck search interface has been released! Watch the ECIR 2020 demo, and try the system out online
  • (Apr) Congratulations to Mahshad Mahdavi, who has accepted a research position at Apple in Cupertino. She will join Apple in mid-August, after finishing her PhD dissertation.
  • (Apr) Congratulations to Yancarlos Diaz, who received the Outstanding 5th-year Scholarship from the RIT computer science department!
  • (Apr) Mahshad Mahdavi's paper on CNN+attention-based formula recognition has been accepted for publication at the CVPR Workshop on Text and Documents in the Deep Learning Era.
  • (Apr) ECIR 2020: in addition to the MathDeck demo, Wei Zhong presented his paper on accelerating formula search, and the ARQMath QA/formula retrieval task was presented in the CLEF 2020 session. Thanks to Behrooz Mansouri for all his good/hard work on preparing the data, evaluation tools, and baselines for the task.

2019


  • (Dec) We will have three papers at ECIR 2020: one full paper on accelerated formula search, a demo paper on our new MathDeck interface, and a short paper on the ARQMath lab at CLEF 2020. Congratulations to everyone on the project who contributed to these papers (see the Publications page).
  • (Dec) The dprl lab welcomes Jessica Diehl who will work on formula indexing in Spring 2020, and is happy to have Yancarlos Diaz returning to work on the MathDeck interface in Spring 2020 as well.
  • (Dec) We are looking for a new PhD student to start in Fall 2020 - please see the "Jobs" page.
  • (Nov) The ARQMath lab has begun! Please visit the lab web page online, and join the ARQMath forum if you are interested in the lab discussions. 
  • (Oct) Congratulations to Abishai Dmello, who has successfully defended his thesis, Representing Mathematical Concepts Associated with Formulas using Math Entity Cards.
  • (Oct) Source code for our formula extraction (ScanSSD), formula recognition (LPGA), and Tangent-CFT formula embedding model are now available through the Software page.
  • (Oct) Demonstration videos of the search interface prototype are now available. These videos demonstrate formula input using handwriting, LaTeX, and images, as well as autocompletion and symbol/formula descriptions provided by our new math entity cards.
  • (Oct) Richard Zanibbi and Abishai Dmello visited the University of Waterloo in Canada, and gave a talk about progress to date for the MathSeer project. 
  • (Sep) Richard Zanibbi gave a talk about MathSeer at RMIT in Melbourne, Australia. 
  • (Aug) Congratulations to Parag Mali, who has successfully defended his MSc thesis, Scanning Single Shot Detector for Math in Document Images.
  • (Aug) Puneeth Kukkadapu has accepted a new job at ACV Auctions, where he will be doing work on handwriting recognition.
  • (Aug) We welcome Jennifer Liu to the project. Jennifer will be working full-time as a research programmer at RIT during Fall 2019.
  • (June) The CROHME+TFD 2019 handwritten formula recognition and typeset formula detection competition has completed. Congratulations to USTC-iFLYTEK (handwritten formula recognition) and Samsung Research (detection) on winning!
  • (June) Behrooz Mansouri's JCDL 2019 paper, Characterizing Searches for Mathematical Concepts was nominated for a Best Paper award.
  • (May) We welcome Thomas Lazore to the project. Thomas will be working as an intern during Summer 2019. Thomas is currently a student at Monroe Community College in Rochester.
  • (April) Congratulations to Wei Zhong, who received the Best Applications Paper Award at ECIR 2019! His paper. "Structural Similarity Search for Formulas using Leaf-Root Paths in Operator Subtrees" can be found here.
  • (April) Congratulations to Puneeth Kukkadapu, who has received the TSO Logic Master's Student Scholarship from the Department of Computer Science.
  • (April) A new version of Tangent-v for formula search in images has been released. This system (along with SymbolScraper, below) is being presented at ECIR 2019 this month. Search results from the paper are included in the package.
  • (April) Our SymbolScraper tool for extracting precise symbol locations and idenitries in born-digital PDF files has been released. The system was first developed by Ritvik Joshi, and then refined by Parag Mali, Puneeth Kukadapu and Mahshad Mahdavi.
  • (Mar) Congratulations to Parag Mali, who will begin working with the Microsoft AI and Ink group in Seattle this fall.
  • (Mar) Prof. Zanibbi will be teaching a new Information Retrieval course in the Computer Science Department at RIT in the fall.
  • (Feb) The new Approach0 formula search engine by Wei Zhong has been released. 
  • (Jan) The CROHME + TFD 2019 handwritten formula recognition and typeset formula detection competition being held as part of ICDAR 2019 is underway. Mahshad Mahdavi is the lead organizer. Please consider participating!
  • (Jan) Congratulations to Mahshad Mahdavi, who will be interning at Apple in Cupertino this summer.
  • (Jan) We welcome Gavin Nishizawa and Yancarlos Diaz, who will be working as research programmers during the spring.

2018

  • (Sep) We welcome Shaurya Rohatgi to the project. Shaurya has been doing work on auto-completion for search queries containing math since the Spring, and will extend this work for his PhD.
  • (Aug) We welcome Behrooz Mansouri to the project, who will be working on math-aware search engines for his PhD research. He is being co-advised by Profs. Zanibbi and Oard.
  • (Aug) Kenny Davila and Prof. Zanibbi received the Best Paper award at the ICFHR 2018 conference for their paper, "Visual Search Engine for Handwritten and Typeset Math in Lecture Videos and LaTeX Notes.
  • (Aug) Mahshad Mahdavi and Prof. Zanibbi attend the ICFHR 2018 conference in Niagara Falls, NY. Prof. Zanibbi is a General Co-Chair for the conference.
  • (July) Wei Zhong and Prof. Zanibbi attend SIGIR 2018, held in Ann Arbor, Michigan. A poster by Behrooz Mansouri was presented by Prof. Zanibbi at the conference.
  • (July) Tangent-v, Kenny Davila's visual graphics search engine has been released. Kenny has succesfully applied the system to recognizing formulas in raster images, formulas in PDF images, and even searching lecture videos for formulas in rendered LaTeX images.
  • (June) The RIT and PennState groups met for a couple of days at PennState. This was the first face-to-face meeting between the groups. We plan to hold our next face-to-face meeting at RIT next summer.
  • (May) Congratulations to Ritvik Joshi, who has accepted a position with Petuum Corporation in Pittsburgh.
  • (Feb) Congratulations to Puneeth Kukkadapu and Parag Mali, who have secured internships at Apple in Cupertino this summer.
  • (Jan) We welcome Puneeth Kukkadapu and Parag Mali to the project, who will be working as part-time Research Assistants for the project during the spring.

2017

  • (Dec) The NSF Foundation has funded a companion grant to our Sloan project! Congratulations to Profs. Zanibbi, Agarwal, Oard, and Giles.
  • (Dec) Congratulations to Xuan Huang, who has secured a summer internship at Facebook.
  • (Nov) Prof. Zanibbi attends ICDAR 2017 and the GREC 2017 workshop in Kyoto, Japan.
  • (Sep) The first meeting of the Advisory Board and project leaders was held. 
  • (Aug) We give a warm welcome to Mahshad Mahdavi (PhD student, Imaging Science) and Wei Zhong (PhD Student, Computing and Information Sciences) who will be joining the project at RIT, working on math notation recognition and retrieval, respectively. We also welcome MSc students Ritvik Joshi, Rahul Dashora, and Xuan Huang. 
  • (June) The Sloan Foundation has funded our math-aware search project! 

MathSeer is supported by the Alfred P. Sloan Foundation and the National Science Foundation