HDR Video Reconstruction: A Coarse-to-fine Network and A Real-world Benchmark Dataset

Static Scenes

For static scenes, we captured 49 two-exposure and 48 three-exposure sequences, each with 15 − 20 frames. The ground-truth HDR frames for static scenes were generated by merging multi-exposure images. We first averaged images having the same exposure to reduce noise, and then merged multi-exposure images using a weighting function similar to [Deep high dynamic range imaging of dynamic scenes, TOG 2017]. For each scene, we will release 6 − 9 captured frames and the generated HDR frame.

(a) Static Scenes Augmented with Global Motion (2-Exposure)

scene_000 scene_001 scene_002 scene_003 scene_004
scene_005 scene_006 scene_007 scene_008 scene_009
scene_010 scene_011 scene_012 scene_013 scene_014
scene_015 scene_016 scene_017 scene_018 scene_019
scene_020 scene_021 scene_022 scene_023 scene_024
scene_025 scene_026 scene_027 scene_028 scene_029
scene_030 scene_031 scene_032 scene_033 scene_034
scene_035 scene_036 scene_037 scene_038 scene_039
scene_040 scene_041 scene_042 scene_043 scene_044
scene_045 scene_046 scene_047 scene_048






(b) Static Scenes Augmented with Global Motion (3-Exposure)

scene_000 scene_001 scene_002 scene_003 scene_004
scene_005 scene_006 scene_007 scene_008 scene_009
scene_010 scene_011 scene_012 scene_013 scene_014
scene_015 scene_016 scene_017 scene_018 scene_019
scene_020 scene_021 scene_022 scene_023 scene_024
scene_025 scene_026 scene_027 scene_028 scene_029
scene_030 scene_031 scene_032 scene_033 scene_034
scene_035 scene_036 scene_037 scene_038 scene_039
scene_040 scene_041 scene_042 scene_043 scene_044
scene_045 scene_046 scene_047