Amodal Instance Segmentation (AIS) aims to segment the regions of both visible and invisible parts of overlapping objects. The mainstream Mask R-CNN-based methods are unsuitable for thin objects with large overlaps because of their object proposal features with bounding boxes for three reasons. First, capturing the entire shapes of overlapping thin objects is difficult. Second, the bounding boxes of close objects are almost identical. Third, a bounding box contains many objects in most cases. In this paper, we propose a box-free AIS method, Seed-to-Mask, for thin objects with large overlaps. The method specifies a target object using a seed and iteratively extends the segmented region. We have achieved better performance in experiments on artificial data consisting only of thin objects.

The traditional RANSAC samples uniformly in the dataset which is not efficient in the task with rich prior information. This letter proposes GUISAC (Guided Sample Consensus), which samples with the guidance of various prior information. In image matching, GUISAC extracts seed points sets evenly on images based on various prior factors at first, then it incorporates seed points sets into the sampling subset with a growth function, and a new termination criterion is used to decide whether the current best hypothesis is good enough. Finally, experimental results show that the new method GUISAC has a great advantage in time-consuming than other similar RANSAC methods, and without loss of accuracy.

Accurate estimation of the state-of-charge is a crucial need for the battery, which is the most important power source in electric vehicles. To achieve better estimation result, an accurate battery model with optimum parameters is required. In this paper, a gradient-free optimization technique, namely tree seed algorithm (TSA), is utilized to identify specific parameters of the battery model. In order to strengthen the search ability of TSA and obtain more quality results, the original algorithm is improved. On one hand, the DE/rand/2/bin mechanism is employed to maintain the colony diversity, by generating mutant individuals in each time step. On the other hand, the control parameter in the algorithm is adaptively updated during the searching process, to achieve a better balance between the exploitation and exploration capabilities. The battery state-of-charge can be estimated simultaneously by regarding it as one of the parameters. Experiments under different dynamic profiles show that the proposed method can provide reliable and accurate estimation results. The performance of conventional algorithms, such as genetic algorithm and extended Kalman filter, are also compared to demonstrate the superiority of the proposed method in terms of accuracy and robustness.

With the popularity of BitTorrent-like P2P applications, improving its performance has been an active research area. Super-seeding, a special upload policy for the initial seeder, improves the efficiency in producing multiple seeds and reduces the uploading bytes of content initiators, thus being highly expected as a promising solution for improving downloading performance while decreasing uploading cost. However, the overall impacts of super seeding upon BitTorrent performance remain a question and have not been analyzed so far in literature. In this paper, we present an analytical and experimental study over the performance of super-seeding scheme. We attempt to answer the following questions: whether and how much super-seeding saves uploading cost, whether the overall downloading time is decreased by super-seeding, and in which circumstances super-seeding performs worse. Based on the seeding process, our analytical study gives formulas on the new piece distribution time, average downloading time and minimum distribution time for heterogeneous P2P file distribution system with super-seeding. Robust evidence supporting the use (or not) of super-seeding is given based on our worldwide Internet experiments over wide distribution of 250 PlanetLab nodes. With a well-designed experimental scenario, we study the overall download time and upload cost of super seeding scheme under varying seed bandwidth and peer behavior. Results show that super-seeding can save an upload ratio of 20% and does help speeding up swarms in certain modes. Tentative conclusions about the effectiveness of super-seeding and its optimal working circumstances are given with inside mechanism analyzed and negative factor identified. Our work not only provides reference for the potential adoption of super-seeding in BitTorrent and other P2P applications, but also much insights for the tussle of enhancing of Quality of Experience (QoE) and saving cost for a large-scale BitTorrent-like P2P commercial application.

An optimized Built-In Self-Test technology is proposed in this paper. A simplified algebraic model is developed to represent the configurations of single input change circuits. A novel single input change sequence generation technique is designed. It consists of a modified scan shift register, a seed storage array and a series of XOR gates. This circuitry can automatically generate single input change sequences of more unique vectors. Experimental results based on the ISCAS-89 benchmark show that the proposed method can achieve high stuck-at fault coverage with low switching activity during test applications.

In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem.

With great advances in the mobile technology and wireless communications, users expect to be online anytime anywhere. However, due to the high cost of being online, applications are still implemented as partially connected to the server. In many data-intensive mobile client/server frameworks, it is a daunting task to archive and index such a mass volume of complex data that are continuously added to the server when each mobile client gets online. In this paper, we propose a scalable technique called Seeded Clustering that allows us to maintain R-tree indexes by bulk insertion while keeping pace with high data arrival rates. Our approach uses a seed tree, which is copied from the top k levels of a target R-tree, to classify input data objects into clusters. We then build an R-tree for each of the clusters and insert the input R-trees into the target R-tree in bulk one at a time. We present detailed algorithms for the seeded clustering and bulk insertion as well as the results from our extensive experimental study. The experimental results show that the bulk insertion by seeded clustering outperforms the previously known methods in terms of insertion cost and the quality of target R-trees measured by their query performance.

We have proposed and experimentally demonstrated a novel WDM-PON downstream optical link. It is composed of a wavelength-locked FP-LD with a spectrally-sliced FP-LD as an external-injection optical source and operated as directly-modulating in a downstream-traffic transmitter. The downstream transmissions at 622 Mbps and 2.5 Gbps were performed for four channels over 25 km. The proposed WDM-PON downstream transmitter can be expanded up to eight channels by controlling an external-injection optical source of a spectrally-sliced FP-LD. Also, the transmitter has facility of multi-channel selection by controlling temperature. We verified the potential of the transmitter in WDM-PON optical link.

This letter presents a practical approach for high-quality built-in test using a test pattern generator called neighborhood pattern generator (NPG). NPG is practical mainly because its structure is independent of circuit under test and it can realize high fault coverage not only for stuck-at faults but also for transition faults. Some techniques are also proposed for further improvement in practical applicability of NPG. Experimental results for large industrial circuits illustrate the efficiency of the proposed approach.

Reseeding technique is proposed to improve the fault coverage in pseudo-random testing. However most of previous works on reseeding is based on storing the seeds in an external tester or in a ROM. In this paper we present a built-in reseeding technique for LFSR-based test pattern generation. The proposed structure can run both in pseudorandom mode and in reseeding mode. Besides, our method requires no storage for the seeds since in reseeding mode the seeds can be generated automatically in hardware. In this paper we also propose an efficient grouping algorithm based on simulated annealing to optimize test vector grouping. Experimental results for benchmark circuits indicate the superiority of our technique against other reseeding methods with respect to test length and area overhead. Moreover, since the theoretical properties of LFSRs are preserved, our method could be beneficially used in conjunction with any other techniques proposed so far.

We propose a technique of selecting seeds for the LFSR-based test pattern generators that are used in VLSI BISTs. By setting the computed seed as an initial value, target fault coverage, for example 100%, can be accomplished with minimum test length. We can also maximize fault coverage for a given test length. Our method can be used for both test-per-clock and test-per-scan BISTs. The procedure is based on vector representations over GF(2m), where m is the number of LFSR stages. The results indicate that test lengths derived through selected seeds are about sixty percent shorter than those derived by simple seeds, i.e. 0001, for a given fault coverage. We also show that seeds obtained through this technique accomplish higher fault coverage than the conventional selection procedure. In terms of the c7552 benchmark, taking a test-per-scan architecture with a 20-bit LFSR as an example, the number of undetected faults can be decreased from 304 to 227 for 10,000 LFSR patterns using our proposed technique.

Seeding substrates with diamond nanocrystals has been considered to be a promising nondestructive pretreatment method for growth of diamond films. However, its application is strongly impeded by the segregation of diamond nanocrystals on substrates. In the present study, we suggest a very simple but effective seeding way ("sandwich" (SW) seeding way) to prevent nanocrystals from segregation. By the SW seeding way, the diamond nanocrystals can be nearly uniformly dispersed on Si substrates with the areal density of the order of 108cm-2. On the nano-seeded Si substrates the continuous and homogeneous diamond films can be successfully fabricated using a microwave plasma enhanced chemical-vapor-deposition (MPECVD) equipment. The cross-sectional transmission electron microscopy (TEM) images reveal that compare with the diamond films grown on the Si substrates pretreated by the conventional scratching method, the films deposited on the nano-seeded Si substrates present a much flatter interfacial structure, suggesting that the SW seeding way can effectively reduce the interface coarseness.

The effect of NiP as a seed layer for the [Co/Pd]n multilayer perpendicular recording media was studied. It was found that a thin layer of 2 nm NiP inserted between the FeCoC soft magnetic underlayer and the [Co/Pd]n recording layer improved the magnetic properties such as coercivity, squareness and nucleation field. These improvements may be due to the enhanced grain isolation promoted by the NiP seed layer, as well as the lower surface roughness of the NiP seed layer. Read/write test using Guzik spin stand with a ring-type head showed a D50 value 220 kFCI in the roll-off curve. The magnetic transitions recorded up to 390 kFCI for [Co/Pd]n media with the NiP seed layer can be observed clearly with MFM.

This paper presents a new deterministic built-in test scheme using a neighborhood pattern generator (NPG) to guarantee complete fault efficiency with small test-data storage. The NPG as a decoding logic generates both a parent pattern and deterministic child patterns within a small Hamming distance from the parent pattern. A set of test cubes is encoded as a set of seeds for the NPG. The proposed method is practically acceptable because no impact on a circuit under test is required and the design of the NPG does not require the results of test generation. We also describe an efficient seed generation method for the NPG. Experimental results for benchmark circuits demonstrate that the proposed method can significantly reduce the storage requirements when compared with other deterministic built-in test methods.

Ferromagnetic Co77Cr20Ta3 layers were deposited on a Pt seed layer by a facing targets sputtering apparatus. The Co-Cr-Ta and Pt crystallites revealed better c-axis orientation at a substrate temperature Ts above 200C. Relatively high perpendicular coercivity Hc⊥ of 2.5 kOe was obtained for the bilayered film with a Co-Cr-Ta layer thickness, δCo, of 50 nm deposited at Ts of 250C. Although the Co-Cr-Ta/Pt medium with δCo of 100 nm exhibited lower recording density than a Co-Cr-Ta/Cr longitudinal one, its noise level became small at the high-density recording range. Measurement of the anomalous Hall voltage clarified that the bilayered film with δCo as small as 30 nm revealed larger perpendicular magnetization than the single layer. The Pt seed layer is effective for depositing thin ferromagnetic Co-Cr-Ta layers below 100 nm in thickness.