Home▸AI & Machine Learning▸3D Feature Pyramid Network: Scale Assignment, Top-Down Merging and Noise Averaging for Small Objects

🔺 3D Feature Pyramid Network: Scale Assignment, Top-Down Merging and Noise Averaging for Small Objects

Two Blender-built plate pyramids on a workbench, the backbone and the top-down feature pyramid of a detector: one shared head sees squares of any size, a level rule picks the plate, and merging coarse maps into fine ones raises the signal-to-noise ratio of small objects, checked against pooling, closed forms and brute-force search.

AI & Machine Learning3DAdvanced60 FPS📱 Mobile-adapted🟠 Blender⇄ 2D version
3D Feature Pyramid Network: Scale Assignment, Top-Down Merging and Noise Averaging for Small Objects ↗ Open standalone
⚙ Under the hood

Two Blender-built plate pyramids on a workbench, the backbone and the top-down feature pyramid of a detector: one shared head sees squares of any size, a level rule picks the plate, and merging coarse maps into fine ones raises the signal-to-noise ratio of small objects, checked against pooling, closed forms and brute-force search.

Feature Pyramid NetworkObject DetectionMulti-Scale FeaturesAverage PoolingTop-Down PathwayAnchor IoUSignal-to-Noise RatioComputer Vision

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)