Peterson Lab Can infants (5.5 month-olds) use object repetition to segment objects from their backgrounds? Elizabeth Salvagio, Rebecca L. Gomez & Mary A. Peterson University of Arizona 26.334 Support: MAP NSF BCS 0960529 Contact: Elizabeth Salvagio bsalvag@email.arizona.edu Stimuli & Schematic Background References 1) Brady, MJ & Kersten, D (2003). Bootstrapped learning of novel objects. Journal of Vision, 3, 413-422. 2) Granrud, CE & Yonas A (1984). Infants’ Perception of Pictorially Specified Interposition. Journal of Experimental Child Psychology, 37, 500-511. 3) Granrud, EC, Yonas A & Opland EA (1985). Infants’ senstivity to the depth cue of shading. Perception & Psychophysics, 37, 415-429. 4) Imura, T, Yamaguchi, MK, Kanazawa, S, Shirai, N, Otsuka, Y, Tomonaga, M & Yagi, A (2008). Infants’ senstiviity to shading and line junctions. Vision Research, 48, 1420-1426. Habituation Paradigm with Looking Times 2 Novel Objects (A & B) 8 Backgrounds Pre-test order: AB or BA Post-test order nested: ABAB or BABA Object & Labeling Phrase: Constant Evidence for novel object segmentation Pre-test Habituation “Look at the Modi (Toma). Do you see the Modi (Toma)?” “Look at the Modi (Toma). Do you see the Modi (Toma)?” “Look at the Modi (Toma). Do you see the Modi (Toma)?” “Look at the Modi (Toma). Do you see the Modi (Toma)?” Post-test “Look at the Modi (Toma). Do you see the Modi (Toma)?” A “Look at the Modi (Toma). Do you see the Modi (Toma)?” B Experiment 2 3D novel objects & variable 2D backgrounds Experiment in progress Does repetition of 3D object on variable 3D backgrounds support segmentation? Conclusions 3D novel objects & constant 2D background Segmentation? YES Recognition, therefore Segmentation when: 2D backgrounds change across trials But is a variable background necessary? Or do shading & interposition support segmentation? 7mos but not 5mos sensitive to these depth cues 2,3,4 Segmentation? NO Some adults can learn to recognize a novel object when: Viewed repeatedly against variable static backgrounds Texture and shading of object and background identical 1 Experiment 1 Segmentation? NO 5.5mos can use object repetition to segment only when ground varies & not when ground constant Emergence of past experience cue mere repetition of object on changing ground leads to segmentation First demonstration that infants can use this cue Unlike adults they use this cue without explicit instruction and in a single learning experience How do infants segment scenes? Can repetition of a novel object on variable backgrounds lead to segmentation? BUT! Displays dimmer Eliminated > 2x N as in previous Experiments for fussiness & Failure to habituate No Recognition/Segmentation when: 2D Background constant across trials No conclusions possible regarding 3D backgrounds N=16 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60 Habituation Object Other Object Average Looking Time (s) Pre-test Post-test * 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60 Habituation Object Other Object Average Looking Time (s) Pre-test Post-test N=11 n.s. 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60 2.80 3.00 3.20 3.40 Habituation Object Other Object Habituation Object Other Object Variable Ground Constant Ground Average Looking Time (s) Pre-test Post-test n.s. n.s. N=13 N=10