Jared Mejia

Papers

Total Citations

H-Index

About

Jared Mejia is pioneering the integration of multimodal perception for robotic manipulation, with a focus on how touch and sound can enhance machine learning. His most-cited work, "Hearing Touch: Audio-Visual Pretraining for Contact-Rich Manipulation" (2024, 9 citations), challenges the dominant paradigm of visual-only pretraining by demonstrating that audio signals—captured from physical contact—can be leveraged to build richer, more robust representations for robots. This contribution is particularly significant for contact-rich tasks, where tactile feedback is critical but often overlooked. By showing that pretraining on sound can improve manipulation performance without requiring extensive real-world data, Mejia opens new avenues for scalable robot learning. His work bridges sensory gaps, making robots more adaptable to unstructured environments. As an emerging researcher, Mejia is already shaping conversations around multimodal pretraining, and his innovative approach promises to influence future designs in embodied AI and dexterous robotics.