Person:

Zickler, Todd

We propose an approach for cross-view action recognition by way of ‘virtual views’ that connect the action descriptors extracted from one (source) view to those extracted from another (target) view. Each virtual view is associated with a linear transformation of the action descriptor, and the sequence of transformations arising from the sequence of virtual views aims at bridging the source and target views while preserving discrimination among action categories. Our approach is capable of operating without access to labeled action samples in the target view and without access to corresponding action instances in the two views, and it also naturally incorporate and exploit corresponding instances or partial labeling in the target view when they are available. The proposed approach achieves improved or competitive performance relative to existing methods when instance correspondences or target labels are available, and it goes beyond the capabilities of these methods by providing some level of discrimination even when neither correspondences nor target labels exist.

Consumer digital cameras use tone-mapping to produce compact, narrow-gamut images that are nonetheless visually pleasing. In doing so, they discard or distort substantial radiometric signal that could otherwise be used for computer vision. Existing methods attempt to undo these effects through deterministic maps that de-render the reported narrow-gamut colors back to their original wide-gamut sensor measurements. Deterministic approaches are unreliable, however, because the reverse narrow-to-wide mapping is one-to-many and has inherent uncertainty. Our solution is to use probabilistic maps, providing uncertainty estimates useful to many applications. We use a non-parametric Bayesian regression technique - local Gaussian process regression - to learn for each pixel's narrow-gamut color a probability distribution over the scene colors that could have created it. Using a variety of consumer cameras we show that these distributions, once learned from training data, are effective in simple probabilistic adaptations of two popular applications: multi-exposure imaging and photometric stereo. Our results on these applications are better than those of corresponding deterministic approaches, especially for saturated and out-of-gamut colors.

We propose modifying the aperture of a conventional color camera so that the effective aperture size for one color channel is smaller than that for the other two. This produces an image where different color channels have different depths-of-field, and from this we can computationally recover scene depth, reconstruct an all-focus image and achieve synthetic re-focusing, all from a single shot. These capabilities are enabled by a spatio-spectral image model that encodes the statistical relationship between gradient profiles across color channels. This approach substantially improves depth accuracy over alternative single-shot coded-aperture designs, and since it avoids introducing additional spatial distortions and is light efficient, it allows high-quality deblurring and lower exposure times. We demonstrate these benefits with comparisons on synthetic data, as well as results on images captured with a prototype lens.

The color of a scene recorded by a trichromatic sensor varies with the spectral distribution of the illuminant. For recognition and many other applications, we seek to process these measurements to obtain a color representation that is unaffected by illumination changes. Achieving such color constancy is an ill-posed problem because both the spectral distribution of the illuminant and the scene reflectance are unknown. For the most part, methods have approached this problem by leveraging the statistics of individual pixel measurements, independent from their spatial contexts. In this work, we show that the strong spatial correlations that exist between measurements at neighboring image points encode useful information about the illuminant and should not be ignored. We develop a method to encode these correlations in a statistical model and exploit them for color constancy. The method is computationally efficient, allows for the incorporation of prior information about natural illuminants, and performs well when evaluated on a large database of natural images.

We consider the problem of finding distinctive social interactions involving groups of agents embedded in larger social gatherings. Given a pre-defined gallery of short exemplar interaction videos, and a long input video of a large gathering (with approximately-tracked agents), we identify within the gathering small sub-groups of agents exhibiting social interactions that resemble those in the exemplars. The participants of each detected group interaction are localized in space; the extent of their interaction is localized in time; and when the gallery of exemplars is annotated with group-interaction categories, each detected interaction is classified into one of the pre-defined categories. Our approach represents group behaviors by dichotomous collections of descriptors for (a) individual actions, and (b) pair-wise interactions; and it includes efficient algorithms for optimally distinguishing participants from by-standers in every temporal unit and for temporally localizing the extent of the group interaction. Most importantly, the method is generic and can be applied whenever numerous interacting agents can be approximately tracked over time. We evaluate the approach using three different video collections, two that involve humans and one that involves mice.

In “Shape From Specular Flow: Is One Flow Enough?” (Vasilyev, et al., 2011 [5]) we show that mirror shape can often be reconstructed from the observation of a single specular flow. In this technical report we provide additional details that, due to space constraints, could not be included in the paper. First we provide a derivation of the linear system for the reflection field derivative in the direction orthogonal to the flow, ˆry. Second, we derive an expression for the determinant of this system which is independent of coordinate system. Third, we show that the sphere is reconstructable whenever the scene rotation is neither on the equator nor parallel to the view direction. Finally we provide additional details for the outline of the proof that reconstructability is a generic property and for our numerical investigation of the dimensionality of the variety described by the “bad” conditions.

In an extended image sequence of an outdoor scene, one observes changes in color induced by variations in the spectral composition of daylight. This paper proposes a model for these temporal color changes and explores its use for the analysis of outdoor scenes from time-lapse video data. We show that the time-varying changes in direct sunlight and ambient skylight can be recovered with this model, and that an image sequence can be decomposed into two corresponding components. The decomposition provides access to both radiometric and geometric information about a scene, and we demonstrate how this can be exploited for a variety of visual tasks, including color-constancy, background subtraction, shadow detection, scene reconstruction, and camera geo-location.

Motivated by perceptual principles, we derive a new color space in which the associated metric approximates perceived distances and color displacements capture relationships that are robust to spectral changes in illumination. The resulting color space can be used with existing image processing algorithms with little or no change to the methods.

Estimating the color of a scene illuminant often plays a central role in computational color constancy. While this problem has received significant attention, the methods that exist do not maximally leverage spatial dependencies between pixels. Indeed, most methods treat the observed color (or its spatial derivative) at each pixel independently of its neighbors. We propose an alternative approach to illuminant estimation-one that employs an explicit statistical model to capture the spatial dependencies between pixels induced by the surfaces they observe. The parameters of this model are estimated from a training set of natural images captured under canonical illumination, and for a new image, an appropriate transform is found such that the corrected image best fits our model.