Recent advances in transient imaging: A computer graphics and vision perspective

Abstract

Figure 1: Left: In 1964, Harold Edgerton captured the iconic Bullet Through Apple image ( c MIT Museum). The bullet traveled at about 850 m/s, which translated into an exposure of approximately 4-10 millionth of a second. Right: Almost 50 years later, the femto-photography technique was introduced [Velten et al. 2013], capable of capturing light in motion, with an effective exposure time of one trillionth of a second. The large split image is a composite of the three complete frames shown in the insets. The complete videos of this and other scenes can be downloaded from AbstractTransient imaging has recently made a huge impact in the computer graphics and computer vision fields. By capturing, reconstructing, or simulating light transport at extreme temporal resolutions, researchers have proposed novel techniques to show movies of light in motion, see around corners, detect objects in highly-scattering media, or infer material properties from a distance, to name a few. The key idea is to leverage the wealth of information in the temporal domain at the pico or nanosecond resolution, information usually lost during the capture-time temporal integration. This paper presents recent advances in this field of transient imaging from a graphics and vision perspective, including capture techniques, analysis, applications and simulation.

Tackling 3D ToF Artifacts Through Learning and the FLAT Dataset

“…We leverage the availability of computational power and advances in transient rendering [20,10] to synthesize a training dataset sufficiently large and diverse to explore a much larger class of learned models. In addition to physically-accurate MPI effects, the dataset provides realistic shot noise, supports augmentation with approximate motion models, and allows for efficient generation of raw measurements from AMCW ToF sensors with arbitrary modulation signals.…”

“…Depth error for scenes from the FLAT dataset, corrupted by shot noise and MPI and reconstructed by DToF[10], Phasor Imaging[18], LF2[1] and MRM, in cm. Errors are computed only in the unanbiguous reconstruction range of Phasor Imaging,[1.5m, 5m]; no mask is used to remove unreliable pixels.…”

Bidirectional Rendering of Vector Light Transport

“…The emergence of transient imaging [JMMG17] have brought an increased interest on simulating transient light transport in graphics and vision [Jar12, PBSC14, ADY*17]. Jarabo et al .…”

A trillion frames per second: the techniques and applications of light-in-flight photography

“…Light-in-Flight is sometimes also referred to as 'transient imaging', in particular in the context of computer graphics and vision. Jarabo et al have prepared a recent review of the computational methods adopted for LiF or transient imaging [1] and Bhandari et al provide a tutorial overview for the more general problem of time-of-flight imaging [2]. In this review we will focus attention more on the available illumination/detection technologies and applications of LiF, with only a brief overview of some of the computational retrieval techniques.…”