<p>The ability to visually recognize objects despite differences in orientation would be advantageous for fish because they see objects from many viewpoints as they navigate their three-dimensional aquatic environment. We tested the ability of goldfish to recognize 3D chromatic and achromatic stimuli from four aspect angles in three rotation planes using a two-alternative forced-choice task. The fish were trained to discriminate between plastic models of a frog and turtle at 0°, then tested with the same objects at 0°, 90°, 180°, and 270°. In Experiments 1 and 2, the stimuli were presented in color, whereas in Experiment 3, the same stimuli were painted black. In Experiment 1, the fish performed significantly better than chance at all aspect angles (0°, ±90°, 180°) and in all three rotation planes. The goldfish displayed viewpoint-invariant performance in the picture plane, but showed enhanced performance at 0° for the two depth plane rotations, which suggests some viewpoint-dependent processes. In Experiment 2, performance accuracy was high regardless of whether the reinforced stimulus (S+) or the non-reinforced stimulus (S−) was rotated in the picture plane. In Experiment 3, two of four fish were successful in recognizing rotated achromatic stimuli. These results, taken together with other studies, suggest that goldfish more easily achieve visual object constancy when the stimuli contain surface features (color, texture, shading).</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Visual perception of rotated chromatic and achromatic 3D stimuli in goldfish (Carassius auratus)

  • Caroline M. DeLong,
  • Kathryn Gardner,
  • Jessica J. Wegman,
  • Kenneth Tyler Wilcox

摘要

The ability to visually recognize objects despite differences in orientation would be advantageous for fish because they see objects from many viewpoints as they navigate their three-dimensional aquatic environment. We tested the ability of goldfish to recognize 3D chromatic and achromatic stimuli from four aspect angles in three rotation planes using a two-alternative forced-choice task. The fish were trained to discriminate between plastic models of a frog and turtle at 0°, then tested with the same objects at 0°, 90°, 180°, and 270°. In Experiments 1 and 2, the stimuli were presented in color, whereas in Experiment 3, the same stimuli were painted black. In Experiment 1, the fish performed significantly better than chance at all aspect angles (0°, ±90°, 180°) and in all three rotation planes. The goldfish displayed viewpoint-invariant performance in the picture plane, but showed enhanced performance at 0° for the two depth plane rotations, which suggests some viewpoint-dependent processes. In Experiment 2, performance accuracy was high regardless of whether the reinforced stimulus (S+) or the non-reinforced stimulus (S−) was rotated in the picture plane. In Experiment 3, two of four fish were successful in recognizing rotated achromatic stimuli. These results, taken together with other studies, suggest that goldfish more easily achieve visual object constancy when the stimuli contain surface features (color, texture, shading).