Ilker Yildirim
Assistant ProfessorCards
About
Research
Publications
2024
Temporal Segmentation and “Look Ahead” Simulation: Physical Events Structure Visual Perception of Intuitive Physics
Yates T, Yasuda S, Yildirim I. Temporal Segmentation and “Look Ahead” Simulation: Physical Events Structure Visual Perception of Intuitive Physics. Journal Of Experimental Psychology Human Perception & Performance 2024, 50: 859-874. PMID: 38900527, DOI: 10.1037/xhp0001218.Peer-Reviewed Original ResearchAttentional consequencesIntuitive physicsEvent representationPerformance-based tasksVisual perceptionExperiment 1Experiment 2Event boundariesTemporal structurePhysical eventsSensory inputVisual inputObject interactionsTemporal segmentationPerceptionPhysical worldPhysical objectsPrediction partPhysicsMindfulnessImages with harder-to-reconstruct visual representations leave stronger memory traces
Lin Q, Li Z, Lafferty J, Yildirim I. Images with harder-to-reconstruct visual representations leave stronger memory traces. Nature Human Behaviour 2024, 8: 1309-1320. PMID: 38740989, DOI: 10.1038/s41562-024-01870-3.Peer-Reviewed Original ResearchReconstruction errorFeature embeddingScene imagesAdaptive modulationLevels of processing theoryReconstruction residualsStronger memory tracesPerception interfaceInterface perceptionMemory accuracyVisual representationInfluence memoryPerceptual processingMemory durabilityMemory tracesResponse latencyMemoryImagesErrorEmbeddingDatasetArchitectureRetrievalIntentional selectionReconstructionThree-Dimensional Reconstruction Pre-Training as a Prior to Improve Robustness to Adversarial Attacks and Spurious Correlation
Yamada Y, Zhang F, Kluger Y, Yildirim I. Three-Dimensional Reconstruction Pre-Training as a Prior to Improve Robustness to Adversarial Attacks and Spurious Correlation. Entropy 2024, 26: 258. PMID: 38539769, PMCID: PMC10968904, DOI: 10.3390/e26030258.Peer-Reviewed Original ResearchAdversarial trainingPre-trainingAdversarial attacksAdversarial robustnessRobustness of image classifiersModel of human visionComputational model of human visionAdversarial examplesImage classifierWeight initializationDataset settingData augmentationBackground textureSpurious correlationsHuman visionModels of visionImprove robustnessDatasetRobustnessImage formationAttacksComputational modelTrainingShapeNetAdversaryFrom task structures to world models: what do LLMs know?
Yildirim I, Paul L. From task structures to world models: what do LLMs know? Trends In Cognitive Sciences 2024, 28: 404-415. PMID: 38443199, DOI: 10.1016/j.tics.2024.02.008.Peer-Reviewed Original Research
2023
Perception of 3D shape integrates intuitive physics and analysis-by-synthesis
Yildirim I, Siegel M, Soltani A, Ray Chaudhuri S, Tenenbaum J. Perception of 3D shape integrates intuitive physics and analysis-by-synthesis. Nature Human Behaviour 2023, 8: 320-335. PMID: 37996497, DOI: 10.1038/s41562-023-01759-7.Peer-Reviewed Original ResearchThree-dimensional shape perceptionShape perceptionIntuitive physicsHuman shape perceptionDeep neural network modelMachine vision systemHuman performanceDifficult discriminationNeural network modelHuman observersSurface cuesVision systemPerceptionAdequate accountGenerative modelNetwork modelResponse timeCuesImage formationObjectsDiscriminationRobustnessObserverModelInference