FireViewer v2.5 · selected cascade models
Four selected native PyTorch models for probable-fire proposals, visible flames, visual anchors and smoke compatibility.
Image Classification • 22.1M • Updated • 8.46kNote Estimate fire-associated smoke compatibility with an abstention interval. No smoke point, source location, mask or physical aerosol density is inferred.
fireviewer/rfdetr-medium-probable-fire-v2.5
Object Detection • 34.6M • Updated • 78Note Generate single-class probable_fire candidate regions from full images. Flame and smoke are merged at this proposal stage. Selected-validation scores are not the external cascade TEST.
fireviewer/rfdetr-medium-visible-flame-v2.5
Object Detection • 34.6M • Updated • 70Note Find visible flames inside probable-fire proposals. Crop specialist; not an independently qualified full-image detector. Tiny-flame localization remains weak in the external diagnostic.
fireviewer/dinov2-small-visible-flame-monopoint-v2.5
22.6M • Updated • 71Note Predict one visual flame anchor for each detected flame box. Trained on pseudo-points. No independent human flame-base accuracy or calibrated point confidence is established.