Multimedia at IBM encompasses research in technology, tools, and usability along with active participation in the development of standards. Our researchers have played leading roles in the definition of international standards for compression and storage of facsimile, DVD, MPEG-1 and MPEG-2 digital video. We also engineered the convergence of the MPEG-4/SMIL textual representation for object-oriented multimedia presentations. Currently, we are leading the development of almost half of the description schemes defined in the MPEG-7 standard. Video chips that we invented were used in equipment that enabled the first HDTV (High Definition Television) newscast, featuring John Glenn's return to space in 1998. The world's first commercial contract for automated parcel mail sorting used our recognition and image processing technology. We have created technology for manipulating photo-realistic panoramas, querying images by content, and bringing the art treasures of the State Hermitage Museum to the World Wide Web. Our current research focuses on the creation, management, protection, and delivery of digital media, and covers a broad range of applications from e-business and entertainment to learning.

Automated Content Identification, Categorization, and Indexing
Through a combination of computer vision, speech recognition, audio time-scale modification, and information retrieval techniques, we are creating systems to visualize video summaries and to enable open-vocabulary speech indexing. We are also investigating joint audio-video segmentation techniques and bridging the semantic gap between low-level information computed from media and the rich semantic interpretation that users associate with the same data. We are pioneering efficient and robust search methods to support similarity matching with fuzzy constraints across the different levels of multimedia content description, including features, structure, and semantics.

By deploying enterprise-wide business media systems, we are gaining new insights into post-production automation for ingesting, transcoding, analyzing, and repurposing. We are contributing to the MPEG-7 standard for enabling interoperable content-based querying of image, video, audio, and multimedia archives.

Management, Distribution, and Protection
Residential broadband rollouts will revolutionize the way people purchase, receive, and experience entertainment. We are developing chips, clients, and servers to enable the set-top box as an interactive services platform. We are also designing XML schemas to flexibly adapt e-commerce interfaces and to give a personalized feel to broadcast content. We have a leading role in interactive television standardization through our activities on a number of standards bodies, including ATSC (Advanced Television Systems Committee), SMPTE (Society for Motion Picture and TV Engineers), MPEG (Moving Pictures Experts Group), and OpenCable.

We are also exploring distribution and cache management algorithms to exploit the isochronous, interactive, and high-value attributes of streaming media. Our content protection systems enable end-to-end security in the presence of untrusted distribution nodes and push the bandwidth and robustness bounds of digital data hiding.

Content Creation
To achieve substantial bandwidth reductions, compression must move beyond today's bandwidth-efficient version of analog audio and video techniques. We are developing content authoring tools to capture and manipulate presentations composed from multiple individual media objects. We are breaking new ground with a novel compression technology especially well suited for wireless clients, real-time transcoding, and wide-scale distribution.

Multimedia-Enhanced Learning
We are exploring the effectiveness of visual concept learning tasks, problem-centered multilearner tutoring environments, and knowledge representation tools for a wide range of purposes in instruction, training, and assessment. From these multimedia interactions, we distill a set of architectures that can be reapplied across modalities and devices, enabling learning to be delivered as an integrated suite of web services.

Please contact Paridhi Verma to obtain copies of the Computer Science Brochure