Downloads · 30 days
0
turhancan97/visbench-surface_normal-dinov2_vitb14
visbench-surface_normal-dinov2_vitb14 is a machine learning model from turhancan97. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for visbench.
A trained probe head, not a backbone. It is the small module VisBench fits on top of frozen dinov2vitb14 features to measure what those features carry.
Downloads · 30 days
0
Access
Public
Updated Aug 7, 2026
Repo size
29.7 KB
Likes
0
Public
Click a slice to open those files.
.pt14.9 KB · 82%
From the Hugging Face model README
dinov2_vitb14A trained probe head, not a backbone. It is the small module VisBench fits
on top of frozen dinov2_vitb14 features to measure what those features carry.
import visbench
from visbench.hub import load_probe_from_hub
backbone = visbench.get_backbone("dinov2_vitb14")
probe = load_probe_from_hub("turhancan97/visbench-surface_normal-dinov2_vitb14", backbone=backbone)
These weights were fitted on features from dinov2_vitb14, taken with
pooling=mean and feature_mode=dense_only. Loading them against
anything else is refused, because the failure is otherwise silent: a head fitted
on one pooling and fed another has the right shapes and produces a plausible,
wrong number.
| backbone | dinov2_vitb14 |
| backbone key | dinov2/dinov2_vitb14/224/7764ea0f912e |
| task | surface_normal (mid_level) |
| pooling | mean (requested mean) |
| feature mode | dense_only |
| layers | None |
Reported scores
| metric | value |
|---|---|
d1 | 0.2104 |
d2 | 0.4730 |
d3 | 0.5979 |
mean | 30.1143 |
median | 24.4548 |
rmse | 37.1147 |
Fitted with:
batch_size: 8epochs: 10head: linearhidden_dim: 512layers: Nonelr: 0.0005normal_source: Noneoptimizer: adamwprotocol: probe3duncertainty_aware: Truewarmup_epochs: 1.5weight_decay: 0.0001Generated by VisBench.