International Institute Of Information Technology Hyderabad
preetirupa.devi@gmail.com
Project Overview
Day-to-day visuomotor tasks like threading a needle or inserting a key into a keyhole rely on our ability to perceive 3D depth. Such actions use a combination of binocular (retinal disparity) and monocular (e.g., motion parallax, linear perspective, relative size, lighting, shadow) depth cues for depth estimations. For individuals with vision loss in one eye (uniocular), the loss of binocular disparity increases reliance on monocular cues. Prior research, my doctoral work, and anecdotal evidence suggest that one-eyed individuals may adapt by leveraging on these cues. For instance, the applicant's earlier studies (Devi et al.,2024) found that uniocular individuals perform comparably to controls with one eye occluded on depth-based visuomotor tasks, albeit below binocular performance. Slight improvements are noted on purely visual depth tasks. Behaviourally, Marotta et al. (1995) observed that uniocular adults tend to move their heads more frequently, likely to gain depth information via motion parallax. Anecdotal cases from sports (e.g., Mansoor Ali Khan “Tiger” Pataudi) and aviation (U.S. Air Force Major Ed Fattmann) shows that individuals can be trained to engage in acute depth-related tasks even with only one eye. Given this, the present grant proposal aims to develop and validate training modules for improving depth-related visuomotor task performance.
We propose to develop and validate training protocols using VR technology to improve depth perception among one-eyed individuals. Here, we propose to use a head mounted virtual reality (HMD-VR) system to develop interactive training modules that will incorporate both visual and visuomotor tasks, leveraging on head and hand tracking features of the VR for data capture. Training will be delivered in three progressive blocks:
• Block 1: Introduction to depth sensations using various monocular cues.
• Block 2: Focused training to estimate slants and curves using isolated cues.
• Block 3: Integration of multiple depth cues, with task difficulty scaled by selectively limiting certain cues to sharpen sensitivity to others.
Pre- and post-training assessments will mirror those used in Devi et al.'s study (Devi et al.,2024, 2025), involving both physical and virtual depth judgment tasks. A pilot study involving 10 uniocular individuals and 10 binocular controls (with monocular occlusion) will evaluate the program’s efficacy. Transfer of training to activities of daily living (e.g., inserting key into a keyhole or fitting a bulb into a bulb holder) will also be examined to determine real-world applicability. Our training paradigm differs from what has been developed in the past in one fundamental way. Previous training modules are meant to enhance 3D depth estimations by enabling the two eyes work together and use binocular depth cues. The proposed training paradigm is to train the depth perception of one-eyed individuals using monocular depth cues.