VisuApps
From pixel to machine perception — image encoding, filters, lighting, depth vision and AI skills live.
Apps
Thousands of oriented, soft colour Gaussians are projected and blended ("splatted") — no ray marching, so it runs in real time. Toggle splats ↔ rendered to see a photoreal, fly-around scene emerge from many anisotropic blobs. What replaced NeRF.
Computer VisionTry it
A 3×3 kernel travels animated over a test image while a magnifier computes the 9 multiplications and the sum of each convolution position live, and the feature map emerges on the right. Eight classic filter presets (Sobel, Laplace, Sharpen …) are freely editable; the "Learned filters" mode shows typical trained first-layer kernels of a CNN — the same mechanics, but learned instead of engineer-designed.
Computer VisionTry it
The six foundational skills of computer vision live in the browser: classification (MobileNet), object detection (COCO-SSD), person segmentation (BodyPix), pose estimation (MoveNet), monocular depth estimation (principle demo) and OCR (Tesseract) — either on your own webcam or on generated logistics scenes, including two challenges on skill matching.
Computer VisionTry it
An interactive machine-vision lab: a metallic glossy test object with embossing, a fine scratch and a dirt spot under six industrial lighting types (ring light, dome light, dark field, bright field, backlight, line light). An inspection camera shows the grayscale test image live, and a pixel-based visibility score makes it tangible that lighting decides the feasibility of the inspection task.
Computer VisionTry it
A scene as a function instead of a mesh: a tiny network returns colour and density for every point in space. A ray is marched through the volume and the samples are composited into one pixel. Novel viewpoints are computed — including view-dependent shine.
Computer VisionTry it
Many photos become 3D: a ring of cameras triangulates the same feature point from different viewpoints and a point cloud grows. Switch matte / reflective / transparent — on shiny and glassy surfaces the reconstruction visibly falls apart. The motivation for neural rendering.
Computer VisionTry it
From image to number matrix: a procedurally generated harbor sunset image can be zoomed continuously down to single-pixel level, where each cell shows its stored numeric value. Bit depth (1/4/8/24-bit RGB) is switchable live with a threshold slider, R/G/B channel toggles and a computed file size; a click pins a pixel and shows its decimal, hex and binary value.
Computer VisionTry it
An interactive 3D scene on machine depth vision: in stereo mode two cameras with adjustable baseline show the left/right image plus a depth map live, and the disparity d = f·B/z becomes visible at the marked object as a pixel offset. In ToF mode, animated light pulses send echoes back and build the depth map object by object from the travel times.
Computer VisionTry itWant to use these apps in your courses? Let’s find out what works for you.
One form, three matters. We answer personally.
We show heyprof on your own material. Tell us what you teach and we will prepare the conversation around it.