Telekinesis Learning Library
Understand the architectures, mechanisms, and practical use of Physical AI models in the Telekinesis Agentic OS.
Technical tutorials
6 of 6 deep dives
Intermediate · 9 min readLearn how Grounding DINO finds objects described at inference time. Follow image and text features through fusion, query selection, and box decoding, then see where open-vocabulary detection fits in a robot pipeline.
Read tutorial →
Intermediate · 8 min readSee how Qwen-VL turns image patches and a prompt into language and object locations. Learn what the position-aware adapter does, why grounding is generated token by token, and when a general vision-language model is useful in robotics.
Read tutorial →
Basic · 8 min readLearn how a binary mask becomes an ordered outline with outer borders, holes, and hierarchy. Then use contour approximation and geometry to measure parts, inspect defects, and connect image pixels to robot coordinates.
Read tutorial →
Intermediate · 8 min readFollow image features and learned object queries through RF-DETR. See how one-to-one matching produces a prediction set and how architecture search finds useful accuracy–latency choices for real robot hardware.
Read tutorial →
Intermediate · 10 min readFollow a camera image through YOLOX from multi-scale features to final boxes. Learn why it predicts without anchors, separates class and box branches, assigns training examples with SimOTA, and removes duplicates with NMS.
Read tutorial →
Basic · 7 min readSee how Canny edge pixels vote for possible circle centres and radii, why agreement creates peaks in an accumulator, and when this explainable geometric method is better than a learned detector for robot inspection.
Read tutorial →No tutorials match that model or algorithm.