Dev

OCTA — Dev Tier

(3 scenes)

Blind evaluation tier — no ground truth available.

What you get

Measurements (y), ideal forward operator (H), and spec ranges only.

How to use

Apply your pipeline from the Public tier. Use consistency as self-check.

What to submit

Reconstructed signals and corrected spec. Scored server-side.

Parameter Specifications

🔒

True spec hidden — estimate parameters from spec ranges below.

Parameter Spec Range Unit
inter_bscan_time -0.6 – 0.9 ms
bulk_motion -0.24 – 0.36 mm/s
decorrelation_threshold 0.44 – 0.59

Dev Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DiffusionOCT + gradient 0.766 31.91 0.942 0.85 ✓ Certified Zhang et al., NeurIPS 2024
2 OCT-ViT + gradient 0.762 32.18 0.945 0.81 ✓ Certified Tian et al., ICCV 2024
3 SpeckleFormer + gradient 0.750 30.95 0.93 0.84 ✓ Certified Devalla et al., ECCV 2024
4 ScoreOCT + gradient 0.749 31.04 0.931 0.83 ✓ Certified Wei et al., ECCV 2025
5 RetinalFormer + gradient 0.722 28.63 0.893 0.89 ✓ Certified Chen et al., ICCV 2024
6 OCTA-Net + gradient 0.695 28.03 0.881 0.81 ✓ Certified Hybrid U-Net+Transformer, 2023
7 NLM-OCT + gradient 0.673 26.6 0.848 0.84 ✓ Certified Buades et al., Multiscale Model. Simul. 2005
8 TV-Denoising + gradient 0.660 26.15 0.836 0.82 ✓ Certified Rudin et al., Phys. A 1992
9 U-Net-OCT + gradient 0.650 25.41 0.815 0.85 ✓ Certified Ronneberger et al., MICCAI 2015 (OCT variant)
10 Speckle-Lee + gradient 0.623 24.47 0.785 0.82 ✓ Certified Lee, IEEE TGRS 1980
11 Speckle-DenoiseNet + gradient 0.610 23.85 0.763 0.83 ✓ Certified Devalla et al., BOE 2019
12 BM4D + gradient 0.606 24.1 0.772 0.78 ✓ Certified Maggioni et al., IEEE TIP 2013
13 FFT-OCT + gradient 0.558 22.06 0.693 0.8 ✓ Certified Analytical baseline

Visible Data Fields

y H_ideal spec_ranges

Dataset

Format: HDF5
Scenes: 3

Scoring Formula

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%
Back to OCTA