XMR Demo Industrial Foreign Object Detection Lentils

--- pretty_name: XMR Demo — Industrial Foreign-Object Detection (Lentils) — trained pipelines license: apache-2.0 library_name: cuvis-ai pipeline_tag: image-segmentation tags: - hyperspectral - hyperspectral-imaging - anomaly-detection - foreign-object-detection - food-safety - industrial-inspection - dinomaly - dinov2 - cuvis-ai - cu3s - cubert - xmr --- <p align="center"> <img src="https://raw.githubusercontent.com/cubert-hyperspectral/cuvis.sdk/main/branding/logo/banner.png" alt="Cubert Hyperspectral" width="560"/> </p> <p align="center"> <a href="https://docs.cuvis.ai"><img src="https://img.shields.io/badge/Docs-docs.cuvis.ai-0aa?logo=readthedocs&logoColor=white" alt="Cuvis.AI docs"/></a> <a href="https://github.com/cubert-hyperspectral/cuvis-ai"><img src="https://img.shields.io/badge/GitHub-cuvis--ai-24292e?logo=github" alt="Cuvis.AI on GitHub"/></a> <a href="https://huggingface.co/datasets/cubert-gmbh/XMR_Demo_Industrial_Foreign_Object_Detection_Lentils"><img src="https://img.shields.io/badge/Dataset-HuggingFace-ffd21e?logo=huggingface&logoColor=000" alt="Companion dataset"/></a> </p> # Trained Dinomaly anomaly-detection pipelines for industrial lentils Three pre-trained [Cuvis.AI](https://docs.cuvis.ai) pipelines for hyperspectral foreign-object detection on a sliding lentil conveyor. All three use the same architecture (a frozen DINOv2 backbone with a Dinomaly anomaly head and a two-stage gate-then-quantile binary decider); they differ only in **which three bands the channel-selector node feeds the detector**. The companion [dataset repo](https://huggingface.co/datasets/cubert-gmbh/XMR_Demo_Industrial_Foreign_Object_Detection_Lentils) contains the CU3S session and annotations used by the notebook. This page is the model-side reference: pipeline folders, hparams, weights, and the minimum code to run inference. ## Pipelines | Subfolder | Selector node | Wavelengths (nm) | Image threshold | Quantile | Weight (.pt) | |---|---|---|---|---|---| | `dinomaly_rgb_full_pipeline/` | `FixedWavelengthSelector` | (650, 550, 450) | 0.140 | 0.995 | `dinomaly_multifile_rgb_two_stage.pt` | | `dinomaly_cir_full_pipeline/` | `CIRSelector` | NIR=860, R=670, G=560 | 0.140 | 0.995 | `dinomaly_multifile_cir.pt` | | `dinomaly_custom_selector_full_pipeline/` | `FixedWavelengthSelector` | (542, 902, 886) | 0.151 | 0.995 | `dinomaly_multifile_custom_tw...

Source context: 0 downloads · 0 likes · Pipeline image-segmentation · Library cuvis-ai · Repo cubert-gmbh/XMR_Demo_Industrial_Foreign_Object_Detection_Lentils