ml

Using pretrained language models as a supervisory signal for vision embeddings — a multi-stage schedule that fine-tunes intermediate layers rather than only the head.

PyTorch · Deep metric learning · Vision-language models · ResNet-50

Turning Maximum Concept Matching from a zero-shot method into a few-shot one with prompt tuning, so a model can tell when an input is nothing it was trained on.

PyTorch · CLIP · Prompt tuning · Maximum Concept Matching

Read the surroundings from a camera, synthesise a pattern that disappears into them — the machine-learning half of a friend’s thesis on a sheet that could change its own skin.

Python · OpenCV · PyTorch · pix2pix

Two deliberately lopsided models — one that never calls a positive wrongly, one that never misses a positive — and a meta-model that reads both.

Python · XGBoost · TensorFlow · Ensemble methods

Retraining SSD-MobileNet detectors on small custom datasets, and measuring how much of the feature extractor is worth freezing — the answer was the first four blocks.

TensorFlow · Object Detection API · Keras · MobileNet