Abstract:Computer vision tasks, such as motion estimation, depth estimation, object detection, etc., are better suited to light field images with more structural information than traditional 2D monocular images. However, since costly data acquisition instruments are difficult to calibrate, it is always hard to obtain real-world scene light field images. The majority of the datasets for static light field images now available are modest in size and cannot be used in methods such as transformer to fully leverage local and global correlations. Additionally, studies on dynamic situations, such as object tracking and motion estimates based on 4D light field images, have been rare, and we anticipate a superior performance. In this paper, we firstly propose a new static light field dataset that contains up to 50 scenes and takes 8 to 10 perspectives for each scene, with the ground truth including disparities, depths, surface normals, segmentations, and object poses. This dataset is larger scaled compared to current mainstream datasets for depth estimation refinement, and we focus on indoor and some outdoor scenarios. Second, to generate additional optical flow ground truth that indicates 3D motion of objects in addition to the ground truth obtained in static scenes in order to calculate more precise pixel level motion estimation, we released a light field scene flow dataset with dense 3D motion ground truth of pixels, and each scene has 150 frames. Thirdly, by utilizing the DistgDisp and DistgASR, which decouple the angular and spatial domain of the light field, we perform disparity estimation and angular super-resolution to evaluate the performance of our light field dataset. The performance and potential of our dataset in disparity estimation and angular super-resolution have been demonstrated by experimental results.

5D Light Field Synthesis from a Monocular Video

A Learning-Based Method Using Epipolar Geometry for Light Field Depth Estimation

Three Dimensional Reconstruction Using a Lenslet Light Field Camera

Generating Real-Scene Hologram Through Light Field Imaging and Deep Learning

Light-field view synthesis using convolutional block attention module

NeuLF: Efficient Novel View Synthesis with Neural 4D Light Field.

Light field angular super-resolution based on intrinsic and geometric information.

Near Eye Light Field Display Based on Human Visual Features.

Dimensionality Reduction for the Real-Time Light-Field View Synthesis of Kernel-Based Models

Unsupervised Dense Light Field Reconstruction with Occlusion Awareness

Light Field Video Compression and Real Time Rendering

Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelines

Diffusion-based Light Field Synthesis

Light-Field Raw Data Synthesis from RGB-D Images: Pushing to the Extreme

Toward Real-World Light Field Super-Resolution

Stereo-Knowledge Distillation from dpMV to Dual Pixels for Light Field Video Reconstruction

Dense view synthesis for three-dimensional light-field displays based on position-guiding convolutional neural network

Wide-baseline view synthesis for light-field display based on plane-depth-fused sweep volume

A New Parallel Intelligence Based Light Field Dataset for Depth Refinement and Scene Flow Estimation

A Fast View Synthesis Implementation Method for Light Field Applications.

Dense Light Field Reconstruction from Sparse Sampling Using Residual Network