Visual Perspective
Visual perspective, encompassing the understanding and representation of viewpoints, is a burgeoning research area focusing on how humans and machines perceive and interpret visual information from different angles. Current research emphasizes improving the generalization capabilities of models like Vision-Language Models (VLMs) and developing methods to mitigate biases, particularly cultural biases, in image understanding and analysis. This work is crucial for advancing artificial intelligence, particularly in applications requiring nuanced understanding of context and diverse perspectives, such as autonomous driving, human-robot interaction, and healthcare.
Papers
What Languages are Easy to Language-Model? A Perspective from Learning Probabilistic Regular Languages
Nadav Borenstein, Anej Svete, Robin Chan, Josef Valvoda, Franz Nowak, Isabelle Augenstein, Eleanor Chodroff, Ryan Cotterell
The 3D-PC: a benchmark for visual perspective taking in humans and machines
Drew Linsley, Peisen Zhou, Alekh Karkada Ashok, Akash Nagaraj, Gaurav Gaonkar, Francis E Lewis, Zygmunt Pizlo, Thomas Serre