The Information Sciences Institute (ISI) Center for Vision, Image, Speech and Text Analytics (VISTA) is dedicated to driving innovation in multimedia signal processing, computer vision and data analytics.

VISTA combines ISI’s world-renowned expertise in machine-learning and deep-learning techniques to address pressing challenges in security, surveillance and multimedia content analysis. The center’s experienced academic and research staff push the boundaries of academic knowledge in areas spanning biometrics, face recognition, optical coherence tomography, speech, image processing, forensics and multimedia.

The center’s research funding exceeds $80 million from a variety of sources including research grants, contracts, and gifts.  Ongoing research projects include: 

  • BATL (Biometric Authentication with a Timeless Learner): Biometrics systems are vulnerable to well-crafted presentation attacks. Under IARPA’s Odin/Thor program, ISI researchers are creating systems and algorithms resilient to presentation attacks (spoofing) of biometric authentication systems, including the face, iris and fingerprint systems. Partners include the Switzerland-based Idiap Research Institute, TU Darmstadt, TREX Enterprises, and Northrup Grumman. 

  • DiSPARITY (Digital, Semantic and Physical Analysis of Media Integrity): Sponsored by DARPA’s MediFor program, the ISI research team is focused on identifying and characterizing signs of manipulated images, video and metadata. The team includes collaborators from the University of Erlangen-Nuremburg and the University of Naples Frederico II.

  • ELICIT: Optimal decision making requires ready access to disparate sources of structured (e.g., databases) and unstructured (e.g., natural language) information.  ELICIT researchers are developing a framework that integrates concepts of causality, factual knowledge, and meta-reasoning into a model-driven knowledge graph representation that allows decision makers to access relevant knowledge.  Teammates include Rensselaer Polytechnic Institute, CMU, and Lockheed Martin ATL.

  • ELISA (Exploiting Language Information for Situational Awareness): Today's automatic parsers, translators, extractors, and dictionaries cover a tiny fraction of the world's languages. Can we use general knowledge of how language works to extend the reach of natural language tools?  In this project, we develop technology for rapidly constructing information extraction (IE), machine translation (MT), and topic and sentiment processing capabilities for new languages.  Our collaborators are ICSI, Brno University of Technology, University of Pennsylvania, University of Notre Dame, Rensselaer Polytechnic Institute, and Next Century Corp.  This project is sponsored by DARPA's LORELEI program.

  • GAIA (Generating Alternatives for Interpretation and Analysis): As part of the DARPA AIDA program, the GAIA research team is developing techniques and systems to extract and synthesize knowledge from disparate information sources (multi-lingual texts, including texts embedded in images/videos, speech, image, video, meta-data) to generate multiple, alternate hypotheses/interpretations for analyst consumption. Extracted knowledge will be linked across input sources and stored in a shared repository and hypothesis generation algorithms will identify alternative assertions as reported in different contexts.  The team includes collaborators from Rensselaer Polytechnic Institute (Heng Ji), Columbia (Shih-Fu Chang and Kathleen McKeown), and University of Florida (Daisy Wang).

  • GLAIVE (Graphics-based Learning Approach Integrated with Vision Elements):  The GLAIVE team is funded by IARPA’s Janus to develop face recognition technology that help computers recognize individuals “in the wild,” i.e., in photos and videos captured in uncontrolled settings.

  • SARAL (Summarization and domain-Adaptive Retrieval Across Languages): The SARAL team is developing algorithms and technologies that address the critical need for automated solutions that can identify information of interest across a multiplicity of domains, languages, and scenarios.  Specific research tasks include query- and domain-focused summarization of documents, information extraction, information retrieval, and machine translation.  The SARAL team includes researchers from University of Massachusetts - Amherst, Northeastern University, Rensselaer Polytechnic Institute, MIT, Idiap (Switzerland), and the University of Notre Dame.

VISTA researchers are located across the institute’s three sites in Marina del Rey, California; Arlington, Virginia; and Waltham, Massachusetts.