Najib J. Majaj

This study shows that curvature tuning in V2 and V4 of macaque monkeys is adult-like by 30 weeks of age, with position-invariant tuning emerging in V4.

The paper investigates how the ventral stream processes natural images with noise, finding that recognition bandwidth is conserved but noise tolerance increases along the stream.

This study shows that reward timing and sensory cue reliability influence how monkeys allocate foraging effort in a naturalistic task.

The paper investigates how macaque visual cortex represents contrast and naturalistic texture, finding that V4 better decodes texture independently of contrast than V1 or V2.

The paper investigates neural sensitivity to radial frequency patterns in developing macaques, finding that V4 neurons encode global form early, while V1 does not, and that extrastriate areas mature early, suggesting perceptual limits arise from other brain developments.

The paper investigates how naturalistic image structure is processed in macaque visual cortex across development, finding stable neural sensitivity from early ages.

This study tracks how macaque monkeys develop sensitivity to radial frequency patterns, finding improved performance with age and better discrimination of complex shapes.

The paper measures crowding distance and V4 map size in humans to find that V4 size predicts the spacing limit of visual perception.

This paper validates the Bouma law across 50 observers, showing it explains 82-94% of crowding distance variance, with additional factors revealing asymmetries and proposing a standardized Bouma factor b for cross-study comparisons.

EasyEyes enables accurate fixation for online vision testing without gaze tracking, showing consistent crowding thresholds across lab and online settings.

The paper investigates how sensitivity to naturalistic textures depends on high spatial frequencies through experimental manipulations and model observer analysis.

Among dozens of ImageNet-trained networks scored against macaque V4/IT recordings and human behavior, DenseNet-169, CORnet-S and ResNet-101 rank highest, and brain similarity stops tracking accuracy above 70% top-1.

This paper reviews the use of machine learning, particularly deep learning, in biological vision research, discussing its applications, limitations, and future directions.