report form the Lisbon workshop

[Antonio Carlos da Rocha Costa]

Dear Valerie and all,

    Finally I'm at home, again. Sorry for writing to the list on the Lisbon
workshop only now.

    The galleries of pictures of the workshop are all great! It was
wonderful to meet everybody there personally: Richard Gleaves, Ingvild
Roald, Steve Parkhurst, Steven and Bart, Nikos Gramalidis, Guylhem Aznar,
Thomas Hanke, and many others that I haven't contacted before, not to
mention Oliver Streiter and Chiara Vettori, the organizers of the workshop
and editors of the proceedings.

    The workshop was a wonderful meeting. The various approaches to sign
language processing were all presented, specially the two main ones
SignWriting and HamNoSys.

    The invited talks were excelent. Richard gave a perfect historical talk
on SignWriter. Thomas Hanke gave a short tutorial of HamNoSys, and Carol
Neidl and  Robert Lee explained their very important SignStream annotation
system, an important tool for sign language researchers.

    The work going on on sign language synthesis in the Nikos' group (based
on SignWriting) and in the ViSiCast Project (based on HamNoSys) look very
promissing. The ViSiCast project seems to emphasize the realistic production
of signs (so they make heavy use of motion capture), while Nikos' work
emphasizes interactivity, and is great on it.

    Patrice Dalle (Guylhem's advisor) is a well known French researcher in
vision systems. He demonstrated the software he is developing for sign
language recognition from images captured by video cameras. Really
impressive! He is well on the way to produce automatic annotations for
visual aspects of signs and sign phrases.

    The issue of SWML and Unicode for SignWriting arouse in many occasions
in oral presentations and poster sessions. My personal feeling, from all the
discussions (and also from the track that appeared here in the last few
days) is that one should further the Unicode for SignWriting issue, to get a
clearer idea of its potential. SWML and Unicode are not alternatives, they
are complementary, and it is important to understand clearly how we can
profit the best from both.

    Guylhem Aznar will be visting me next week (he is at the moment in Porto
Alegre, with Marianne Stumpf). I hope to talk a little bit more with him
abouth the Unicode stuff.

    Also, it is important to understand that SignWriting and HamNoSys are
not alternatives for the same purpose. They are complementary, useful for
different purposes, and to get a clear picture of such complementarity is
also important, for the best profit of the development of sign language
processing systems.

    In summary, everybody enjoyed the workshop. Looks it will be useful to
organize another one next year. Oliver and I are already talking about that.

    All the best,

    Ant ­? io Carlos