Acknowledgements

Research supported, in part, by HP labs, Palo Alto. The author would like to thank Thad Starner for helping getting the generalized spatial `visual filter' software running on the SGI Reality Engine. Thanks also to: Roz Picard; Neil Gershenfeld; Sandy Pentland; Ted Adelson; Jennifer Healey; and many others, too numerous to mention here, for many interesting and important discussions; and Matt Reynolds, KB2ACE for help in upgrading the outbound ATV system. Thanks to Chris Barnhart for processing hardware, and to HP labs, VirtualVision, Compaq, Ed Gritz, BelTronics, M/A-Com, ShieldWorks, Addenda Electronics, and Virtual Research for lending or donating additional equipment that made the experiments possible.