SignStream project
   News and reviews
   Screenshots
   Documentation
   Product comparison
   Development
Get SignStream
   System requirements
   Registering
   SignStream CDs
Transcripts and videos
Online data repository
Links
   Related projects
   References
Communicate

Currently a new implementation of SignStream (version 3) is in progress.

• Coding schemes—the formal descriptors constraining annotations within a specific domain or organization—are handled independently of transcripts. This simplifies standardization of annotations, while still allowing such schemes to evolve with the work of individual researchers. Transcribers can choose to build on an existing scheme, such as the default ASL scheme included with the program, or to develop their own schemes to suit their own needs.
• Transcripts can be stored in large databases. Searches can be carried out across a wide set of annotations. The search functionality is similar to that of the current version; however, searching is no longer restricted to one database file at a time.
• Transcripts can also be imported/exported using a simple XML format, making it simpler to share data with other XML-aware applications or to write custom applications for data analysis.
• In the current version, video and audio files must be directly referenced by the transcripts, making it difficult to move these files between different media or to substitute alternate files (e.g., using different codecs or time segments). SignStream 3 adds a level of abstraction called Asset Packaging to address these and other media management concerns. This abstraction allows for globally unique assets, making possible distribution of transcripts, including in printed publications, without including the original video or other source data, while providing a simple way for users to obtain the source files separately as needed, on CD, DVD, or via the Web.
• Hierarchical relationships among fields are supported. For example a Hand Shape field could have children fields for Selected Fingers, Aperture, etc.
• Several efforts to simplify data entry will be added, including single-click drop-down menus for sets of known values, custom entry widgets for specific data types or scheme fields, and a user-defined gloss dictionary for entering recurring sets of annotations.

Project History

Recent development (2000- )

has been supported by NSF grant IIS-9912573, CNS-04279883, and HCC-0705749.

SignStream version 3

Java implementation	Iryna Zhuravlova (replaced Jason Boyd as of spring 2004)
Program design, specifications, and testing	Iryna Zhuravlova Robert G. Lee Joan Nash Carol Neidle

Research assistants: Robert G. Lee, Mike Schlang, Joan Nash.

Design and prototyping of the coding scheme module were aided by the work of George Kierstein.

We are grateful to David Greenfield for assistance with release of the SignStream version 2.2.*

Development of the SignStream application (1995-2000)*

was supported by NSF grant IIS-9528985
(C. Neidle, PI; D. MacLaughlin, Postdoctoral Fellow; B. Bahan, Consultant; R.G. Lee, Research Asst.).

Development of capabilities for multiple synchronized video files and audio files (1999-2000)*

was funded by NSF grant EIA-9809340
(C. Neidle, PI; S. Sclaroff, co-PI; D. MacLaughlin, Consultant).

Development of the SignStream prototype (1994-1995)*

was made possible by NSF grant SBR-941056
(C. Neidle, PI; J. Kegl, co-PI; B. Bahan, co-Investigator; D. MacLaughlin, Research Asst.).

* Programming for versions 1 and 2 of SignStream was carried out by David Greenfield in the Department of Humanities Resources at Dartmouth College (Otmar Foelsche, Director).

Previous versions of the Program

SignStream™ Version 2.0. 2000

System requirements: Mac OS (system 8.0 or later; SignStream requires QuickTime™ version 3.0.2 or later).

SignStream™ Version 1.5. 1999

System requirements: Mac OS (system 7.1 or later; SignStream requires QuickTime™ version 2.1 or later).

Personnel

The following people have been involved in various ways in the design and development of SignStream over the years: