Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces Eyettention II, a lightweight deep-learning model designed to generate realistic eye movement scanpaths during reading, capturing fixation location, within-word landing position, and fixation duration. By addressing the limitations of data scarcity in eye-tracking studies, the model is trained end-to-end and demonstrates superior performance in scanpath prediction compared to existing methods, while also aligning with cognitive theories of reading. The results indicate that Eyettention II not only replicates human-like gaze behavior but also has significant implications for enhancing natural language processing applications and psycholinguistic research.
Eyettention II generates human-like eye movement patterns with unprecedented accuracy, even with limited training data.
The way our eyes move while reading provides valuable insights into both the reader's cognitive processes and the properties of the text. In particular, eye-tracking-while-reading data has shown to be highly beneficial in various technological applications, such as enhancing and interpreting language models and inferring a reader's characteristics. However, these applications often rely on large-scale, data-driven models, which demand extensive eye-tracking datasets that are challenging to obtain due to the resource-intensive nature of data collection. To address the challenge of data scarcity, we develop Eyettention II, an end-to-end trained deep-learning model capable of generating realistic scanpaths consisting of a complete set of fixation attributes in chronological order, including fixation location, within-word landing position, and fixation duration. Our model is lightweight, efficiently trainable on limited GPU resources, and closely aligned with cognitive theories. We demonstrate that Eyettention II surpasses state-of-the-art models in scanpath prediction and mirrors human-like gaze behavior by capturing key psycholinguistic phenomena. With its robust performance, Eyettention II holds the potential to drive advancements in natural language processing, facilitate piloting the materials of psycholinguistic experiments, and uncover new insights beyond what is explicitly encoded in theoretical cognitive models.