File Categories

Every file belongs to one of three categories — uploads, models, or results — and each file also carries tags that describe what it is and which module produced it. The category and tags drive both the tree grouping and search.

uploads

Data you bring into the workspace.

• Raw images (raw tag) — source TIFF/MRC stacks, stored under uploads/raw. These are the inputs for denoising, annotation, or segmentation. Data uploaded for inference is stored here too — there is no separate inference folder.

• Annotations (annotation tag) — training masks and labels, stored under uploads/annotations. They must match the dimensions of their raw images.

• Imported models are placed under uploads/imported_models when you upload external model files.

models

Trained models produced by the training modules.

• Each model carries a method tag (segmentation, denoising, or unspecified) plus a type tag (weights, config, or info).

• Segmentation models live under models/segmentation/<id> and denoising models under models/denoising/<id>.

results

Outputs from the processing modules.

• Each result carries a method tag (segmentation, denoising, annotation, or mesh) plus a type tag (data, info, or wip).

• Results are stored per job under results/segmentation/<id>, results/denoising/<id>, results/meshes/<id>, results/preprocess/<id>, results/stitching/<id>, and results/segcleanup/<id>.

Searching by Category

Type a category name (uploads, models, results), a tag, or a tag synonym into the search box to filter files — for example model, mesh, or annotation. See the Search & Filter article for the full keyword list.