Dev

Compressed Ultrafast Photography (CUP) — Dev Tier

(5 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
dmd_encoding_error -0.48 – 0.72 -
streak_sweep_calibration -1.2 – 1.8 -
temporal_spatial_coupling -2.4 – 3.6 -

Dev Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 DAUHST-CUP + gradient 0.806 35.44 0.97 0.82 ✓ Certified Cai et al., NeurIPS 2022 (CUP)
2 STFormer-CUP + gradient 0.792 33.34 0.956 0.88 ✓ Certified Wang et al., CVPR 2022 (CUP)
3 DiffusionCUP + gradient 0.779 33.19 0.954 0.83 ✓ Certified Qiao et al., Nat. Photonics 2020 (updated 2024)
4 PnP-FastDVDnet + gradient 0.748 30.71 0.927 0.85 ✓ Certified Tassano et al., CVPR 2020 (CUP)
5 DeSCI-CUP + gradient 0.705 28.33 0.888 0.83 ✓ Certified Liu et al., IEEE TPAMI 2018 (CUP adapt.)
6 E2E-CNN-CUP + gradient 0.635 24.89 0.799 0.83 ✓ Certified Liang et al., CVPR 2019
7 GAP-TV + gradient 0.483 19.51 0.575 0.79 ✓ Certified Yuan, ICSIP 2016
8 TV-CUP + gradient 0.447 18.17 0.508 0.81 ✓ Certified Gao et al., Nature 2014
9 TwIST-CUP + gradient 0.425 17.43 0.472 0.81 ✓ Certified Bioucas-Dias & Figueiredo, IEEE TIP 2007 (CUP)

Visible Data Fields

y H_ideal spec_ranges

Dataset

Format: HDF5
Scenes: 5

Scoring Formula

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

PSNR: 40% SSIM: 40% Consistency: 20%
Back to Compressed Ultrafast Photography (CUP)