top of page

STEALTH MODE 

pelegh.JPG

Peleg Harel

Co-Founder & CTO, PlaySpear
Computer Vision · Graphics · Engine Development · 

Peleg Harel is a computer scientist, engineer, and a game developer 

His work spans computer vision, real-time graphics, machine learning, 3D rendering, game design, and engine architecture, often connecting disciplines that are traditionally treated as separate.

​

At PlaySpear, Peleg leads the engineering of the studio’s core technology, engine architecture and rendering, gameplay systems, tools, networking, and performance.
 

Snap,

Computer Vision Engineer developing and deploying real-time machine-learning systems on mobile, including skeleton tracking, segmentation, and 9D object tracking used by millions of Snapchat users.

​

3D Reconstruction

Developed systems for reconstructing complete 3D assets from a small number of product images, including novel texturing and projection methods. The pipeline was used to produce thousands of assets for customers including Nike, Zalando, and Famous Footwear.

​​

​

3D Tools & Graphics
Built a 3D modeling application from scratch in C and OpenGL, including custom modeling, rendering, rigging, animation, texturing, UI, and WebGL/WASM technology.

​

Research & Education
M.Sc. in Computer Science from Ben-Gurion University, with research published at CVPR, and B.Sc. in Computer Science from the Technion.

​​

Released Game - Exodus

Peleg developed and released Exodus, a Steam game built from scratch in C and OpenGL. The project included his own 3D engine, rendering systems, math and game logic code, a graphical level editor, and a unique mechanic where the ground changes dynamically in real time.

Patent holder 

​-Improving 3D geometry using AI-predicted depth

-Generating images using text descriptions and 3D objects

-Tracking an object’s position, rotation, and dimensions for AR
-Removing objects from video and reconstructing the background with machine learning

-Recognizing a person’s body and applying visual effects

-Creating an interactive AR wardrobe from real clothing

​​​​​​​​​​​

​

bottom of page