Automatic photo cropping is an important tool for improving visual quality of digital photos without resorting to tedious manual selection. Traditionally, photo cropping is accomplished by determining the best proposal window through visual quality assessment or saliency detection. In essence, the performance of an image cropper highly depends on the ability to correctly rank a number of visually similar proposal windows. Despite the ranking nature of automatic photo cropping, little attention has been paid to learning-to-rank algorithms in tackling such a problem. In this work, we conduct an extensive study on traditional approaches as well as ranking-based croppers trained on various image features. In addition, a new dataset consisting of high quality cropping and pairwise ranking annotations is presented to evaluate the performance of various baselines. The experimental results on the new dataset provide useful insights into the design of better photo cropping algorithms.

A novel redundant mechanism is proposed for embedded memories in this paper. Redundant rows and columns are added into the memory array as in the conventional approaches. However, the redundant rows and columns are divided into row blocks and column blocks, respectively. The reconfiguration is performed at the row (column) block level instead of the conventional row (column) level. Based on the proposed redundant mechanism, we first show that the complexity of the redundancy allocation problem is NP-complete. Thereafter, an extended local repair-most (ELRM) algorithm suitable for built-in implementation is proposed. The complexity of the ELRM algorithm is O(N), where N denotes the number of memory cells. According to the simulation results, the hardware overhead for implementing this algorithm is below 0.17% for a 1024/spl times/2048-b SRAM. Due to the efficient usage of the redundant elements, the manufacturing yield, repair rate, and reliability can be improved significantly.

The authors describe a family of linear systolic arrays for matrix multiplication exhibiting a tradeoff between local storage and the number of processing elements (PEs). The design consists of processors hooked into a linear array with each processor having storage s, 1or=n/s. The input matrices are fed as two speed data streams using fast and slow channels to satisfy the dependencies in the usual matrix multiplication algorithm. While a family of linear arrays have been synthesized for this problem, this technique leads to simpler designs with fewer number of processors and improved delay from input to output. All these designs use the optimal number of processors for local storage in the range 1>

A simple mapping technique is developed to design systolic arrays with limited I/O capability. The technique is used to improve systolic algorithms for some matrix computations on linearly connected arrays of PEs (processor elements) with constant I/O bandwidth. The important features of these designs are modularity with constant hardware in each PE, few control lines, simple data-input/output format, and improved delay time. This technique is extended to design an optimal n square root n-time systolic algorithm for n*n matrix multiplication with O( square root n) I/O bandwidth requirement on a fault-tolerant VLSI model. In this model, the propagation delay is assumed to be proportional to wire length. Fault reconfiguration is achieved by using buffers to bypass faulty PEs, which does not affect the clock rate of the system. The unidirectional flow of control and data assures correctness of the algorithm in the presence of faulty PEs. The design can be implemented on reconfigurable fault-tolerant VLSI arrays using the Diogenes methodology. The present designs are compared to those in the literature and are shown to be superior with respect to I/O format, control, and delay from input to output.>