"... In PAGE 21: ... In the RELATIONAL condition the agent also used the relational strategies described above in an attempt to build a working alliance with subjects, whereas in the NON-RELATIONAL condition this relational behavior was ablated (see Table 1). Table1 . Differences Between RELATIONAL and NON-RELATIONAL Versions of the FitTrack Exercise Advisor Agent ... ..."

"... In PAGE 2: ... In effect, most of these designs were mainly concerned with the implementation of the technology instead of exploring the use of psychomotor parameters in the human system. Table1 a. An overview of recent experiments with new musical instrument designs.... In PAGE 2: ...o.) STEIM, Amsterdam, the Netherlands Web MIDI controller (straingauge based) Michel Waisvisz, STEIM (see Krefeld, 1990) Hands MIDI controller (buttons, ultrasound ranging and mercury switches) Steve O apos;Hearn, Rhode Island School of design, USA, see Moog (1989) Design of a new stick-like MIDI controller, with pressure sensitive buttons) Table1 b. An overview of recent experiments with devices for conducting.... In PAGE 3: ... Table1 d. An overview of experiments with new musical instrument designs involving whole body movements.... In PAGE 3: ...able 1d. An overview of experiments with new musical instrument designs involving whole body movements. Author Motion to sound design Leon Theremin (see Vail, 1993) Capacitively coupled motion detector controls electronic oscillator Chabot (1990) Ultrasound ranging to detect whole body movements and to control MIDI devices Bauer amp; Foss (1992) GAMS: Ultrasound ranging to detect whole body movements and to control MIDI devices Leslie-Ann Coles, USA Bodysuit performance during CHI 92 Coniglio (1992) MIDIDancer bodysuit / Interactor mapping software Yamaha Corp., Japan MIBURI: arm gestures, finger buttons to MIDI translation David Rokeby, Toronto, Canada Very Nervous System: video image processor translates movement into MIDI Fred Kolman, Amsterdam, the Netherlands Video image processor translates movement into MIDI Camurri (1987) Costel opto-electronic human movement tracking system controls a knowledge based computer music system Table1 e. An overview of experiments with new musical instrument designs involving bioelectric signals.... ..."

"... In PAGE 12: ...The 10 data from DataGlove finger bending are ranged from 0 to 90 degree, but the respective DOFs of the hand model have different range as shown in Table3 . Before using the data from DataGlove, we should make data mapping.... ..."

"... In PAGE 4: ... So, we examined the least squares means of the three gloves and we compared them among themselves. Table1 below compares the three gloves ... ..."

"... In PAGE 4: ... 3.1 The Taxonomy of Free Weight Exercises To fulfill the goals, we have identified the most common, representative free weight exercises in the gym environment in Table1 . Each exercise listed here is common and representative in a sense that people frequently use those exercises to train each individual muscle group in the human body [20].... In PAGE 4: ... Finally, in the lower body category, people use deadlift to train quadriceps and standing calf raise to train calves. Table1 also lists the posture required to perform each exercise. The details of how to perform each free weight exercise can be found in reference [21], and some of the exercises will be explained in section 2.... In PAGE 5: ... 3.2 The Accelerometer Glove and the Posture Clip To track the various free weight exercises listed in Table1 , we use accelerometers and incorporated acceleration data with machine learning techniques. Since in free weight exercises people hold and move weights, shown in Fig.... In PAGE 11: ...2 Evaluation of Recognition Goal To figure out whether the algorithms can be applied to a variety of users, say users with different gender, height, weight, level of experience with free weight exercises, etc., we collected data by asking ten subjects to perform the nine exercises listed in Table1 . There were eight male and two female subjects.... ..."

"... In PAGE 9: ...2. Class description Table2 gives a brief description of important classes in SMART, their super classes, the place where they fit in and their functionality. 4.... ..."

"... In PAGE 17: ...g., first-pass reading times: poor fit: 283 ms; good fit: 281 ms, see Table3 ). Thus, the conditions seem largely comparable.... In PAGE 20: ...reated as within-participants and within-items, and the factor thematic fit (i.e., poor fit vs. good fit), which was treated as within-participants, but between-items (see also section Materials). The means of all measures are shown in Table3 ; F-values and significance levels can be found in Table 4. ... ..."

"... In PAGE 12: ... Figure 18 demonstrates robustness to hand occlusion in sequential steps down the left, then right columns. Figure 18: Tracking a face in the presence of passing hand occlusions (sequence: down left then right columns) Table3 shows the results collected from 43 sample points on five tracking runs with active transient hand occlusion of the face. Average perturbation is less than three pixels for X, Y, and Z.... ..."

"... In PAGE 10: ... The differences between computer-assisted and computer-centred determinations of cell locomotion might be due to shortcomings in either system. When computer generated tracks are transcribed by hand, cell speeds, persistences and diffusion coefficients rise artefactually ( Table3 ). This is clearly due to a smoothing effect on the track data.... ..."