6 98 6. Gaze-contingent processing improves mobility performance Figure 6.3: Example virtual indoor environments that were used for the scene recognition task in Experiment 2. The four possible room categories were: bathroom, bedroom, kitchen and living room. and were encouraged to survey and explore their environment with their hands and by looking around. 6.2.6. Experiment 2 With a different set of participants, the second experiment evaluated visual orientation using a scene recognition task and a visual search task. Both tasks were tested in separate trials. In the scene recognition trials, subjects were placed in a randomly selected unfamiliar virtual environment in which they had to identify the room category (bathroom, bedroom, kitchen or living room). In the visual search trials, subjects were situated in a familiarized studio environment, in which they were tasked to sequentially find many prompted target objects within a given time limit of 2 minutes. Virtual environments We used several virtual environments designed by ArchVizPro Studio (Oneiros srl, Italy), obtained from the Unity asset store (Figure6.3. The environments were slightly modified for the experiments: some rooms and objects were re-scaled and some objects were removed or added to create a more realistic and representative representation. In total we used 16 different environments. One environment, the studio environment that was used for the visual search task, was visited frequently to allow familiarization by the subjects. The other 15 environments (three bathrooms, four bedrooms, three kitchens and five living rooms) served as the unfamiliar environments for the scene recognition tasks. The unfamiliar environments were visited maximally twice by the same subject and were only observed with the simulated phosphene vision. The familiarized studio environment was
RkJQdWJsaXNoZXIy MTk4NDMw