Redefining Vision AI with Oculi

Oculi is a US-based startup that aims to revolutionize autonomy with efficient visual intelligence at the edge with their OCULI SPU (Sensor + Processing Unit). As one of the winners of the IC Taiwan Grand Challenge Batch 4 in the AI Core Technologies and Chips, Oculi maximizes efficiency, preserves privacy, and enables real-time intelligence with the IntelliPixel® technology.

Engineering Vision AI for the Machine

Vision AI has often been considered a laggard in AI developments as conventional machine vision sensors are traditionally designed for human viewing, generating massive amounts of redundant image data that must be processed by complex, power‑hungry external hardware. This approach results in slow, inefficient, and fundamentally not scalable data processing that makes it unsuitable for the rapidly growing demands of Vision AI. Most machine vision sensors are seen as approximately 40,000x less efficient than the human eye, with high energy consumption and high latency to generate images that are more compatible with the human eye, with high resolution and high clarity. As vision AI usage expands and grows, this will no longer be a viable option. Oculi’s IntelliPixel® turns each pixel into a smart, programmable sensing unit that processes only essential data, enabling visually smarter devices, machines, and environments.

Introducing IC Taiwan Grand Challenge

To strengthen Taiwan's position as a global semiconductor powerhouse, the National Science and Technology Council (NSTC) organizes the IC Taiwan Grand Challenge. Started in 2024, the competition originally focused purely on innovative IC designs and advanced application solutions. Today, ICTGC also honors teams in the fields of AI Core Technologies and Chips, Smart Mobility, Smart Manufacturing, Smart Medtech, and Sustainability, which can support and contribute to the semiconductor industry and Taiwan’s industry in general. Oculi is selected as one of the winners of ICTGC Batch 4, featuring their IntelliPixel® technology in the AI Core Technologies and Chips category.

Transform Conventional Image Capture and Imaging into Real‑time Perception and Vision AI

Founded in 2019 as a spinout from Johns Hopkins University (JHU), Oculi is a Vision AI-focused startup based in New York who aims to redefine the capabilities of visual sensors through their groundbreaking Software-Defined Vision Sensors™. By integrating intelligence at the pixel level with their IntelliPixel® technology, Oculi enables devices to see and process visual data in real-time, optimizing efficiency and enhancing privacy.

IntelliPixel® was invented specifically for high-speed and low-power vision. It embeds programmable logic directly inside each pixel, enabling on‑sensor event detection and classification at microsecond speeds. Instead of outputting continuous 8‑ or 12‑bit frames to an external ISP, the sensor streams only high‑value metadata—such as motion vectors, gesture codes, or object cues. This event‑driven approach significantly reduces bandwidth and DRAM requirements while cutting power consumption to sub‑milliwatt levels, and inherently protects privacy by avoiding transmission of images the way conventional image sensors do. Its field‑upgradeable firmware enables the deployment of new algorithms and features without respinning silicon, making the platform highly adaptable across products and markets. As a result, it is able to deliver a scalable, AI‑native vision system ideally suited for always‑on smart devices, spaces, and machines.