David R. Martin

Learning to detect natural image boundaries using local brightness, color, and texture cues

The goal of this work is to accurately detect and localize boundaries in natural scenes using local image measurements. We formulate features that respond to characteristic changes in brightness, color, and texture associated with natural boundaries. In order to combine the information from these features in an optimal way, we train a classifier using human labeled images as ground truth. The output of this classifier provides the posterior probability of a boundary at each image location and orientation. We present precision-recall curves showing that the resulting detector significantly outperforms existing approaches. Our two main results are 1) that cue combination can be performed adequately with a simple linear model and 2) that a proper, explicit treatment of texture is required to detect boundaries in natural images.

Local figure–ground cues are valid for natural images

Figure-ground organization refers to the visual perception that a contour separating two regions belongs to one of the regions. Recent studies have found neural correlates of figure-ground assignment in V2 as early as 10-25 ms after response onset, providing strong support for the role of local bottom-up processing. How much information about figure-ground assignment is available from locally computed cues? Using a large collection of natural images, in which neighboring regions were assigned a figure-ground relation by human observers, we quantified the extent to which figural regions locally tend to be smaller, more convex, and lie below ground regions. Our results suggest that these Gestalt cues are ecologically valid, and we quantify their relative power. We have also developed a simple bottom-up computational model of figure-ground assignment that takes image contours as input. Using parameters fit to natural image statistics, the model is capable of matching human-level performance when scene context limited.

EMF1, A Novel Protein Involved in the Control of Shoot Architecture and Flowering in Arabidopsis

Shoot architecture and flowering time in angiosperms depend on the balanced expression of a large number of flowering time and flower meristem identity genes. Loss-of-function mutations in the Arabidopsis EMBRYONIC FLOWER (EMF) genes cause Arabidopsis to eliminate rosette shoot growth and transform the apical meristem from indeterminate to determinate growth by producing a single terminal flower on all nodes. We have identified the EMF1 gene by positional cloning. The deduced polypeptide has no homology with any protein of known function except a putative protein in the rice genome with which EMF1 shares common motifs that include nuclear localization signals, P-loop, and LXXLL elements. Alteration of EMF1 expression in transgenic plants caused progressive changes in flowering time, shoot determinacy, and inflorescence architecture. EMF1 and its related sequence may belong to a new class of proteins that function as transcriptional regulators of phase transition during shoot development.

Intelligent RAM (IRAM): the industrial setting, applications, and architectures

The goal of intelligent RAM (IRAM) is to design a cost-effective computer by designing a processor in a memory fabrication process, instead of in a conventional logic fabrication process, and include memory on-chip. To design a processor in a DRAM process one must learn about the business and culture of the DRAMs, which is quite different from microprocessors. The authors describe some of those differences and their current vision of IRAM applications, architectures, and implementations.

Finding Actions Using Shape Flows

We propose a novel method for action detection based on a new action descriptor called a shape flow that represents both the shape and movement of an object in a holistic and parsimonious manner. We find actions by finding shape flows in a target video that are similar to a template shape flow. Shape flows are largely independent of appearance, and the match cost function that we propose is invariant to scale changes and smooth nonlinear deformation in space and time. The problem of matching shape flows is difficult, however, yielding a large, non-convex, integer program. We propose a novel relaxation method based on successive convexification that converts this hard program into a vastly smaller linear program: By using only those variables that appear on the 4D lower convex hull of the matching cost volume, most of the variables in the linear program may be eliminated. Experiments confirm that the proposed shape flow method can successfully detect complex actions in cluttered video, even with self-occlusion, camera motion, and intra-class variation.

Exploiting On-Chip Memory Bandwidth in the VIRAM Compiler

Many architectural ideas that appear to be useful from a hardware standpoint fail to achieve wide acceptance due to lack of compiler support. In this paper we explore the design of the VIRAM architecture from the perspective of compiler writers, describing some of the code generation problems that arise in VIRAM and their solutions in the VIRAM compiler. VIRAM is a single chip system designed primarily for multimedia. It combines vector processing with mixed logic and DRAM to achieve high performance with relatively low energy, area, and design complexity. The paper focuses on two aspects of the VIRAM compiler and architecture. The first problem is to take advantage of the on-chip bandwidth for memory-intensive applications, including those with non-contiguous or unpredictable memory access patterns. The second problem is to support that kinds of narrow data types that arise in media processing, including processing of 8 and 16-bit data.