We present an algorithm that integrates image co-segmentation into feature matching, and can robustly yield accurate and dense feature correspondences. Inspired by the fact that correct feature correspondences on the same object typically have coherent transformations, we cast the task of feature matching as a density estimation problem in the homography space. Specifically, we project the homographies of correspondence candidates into the parametric Hough space, in which geometric verification of correspondences can be activated by voting. The precision of matching is then boosted. On the other hand, we leverage image co-segmentation, which discovers object boundaries, to determine relevant voters and speed up Hough voting. In addition, correspondence enrichment can be achieved by inferring the concerted homographies that are propagated between the features within the same segments. The recall is hence increased. In our approach, feature matching and image co-segmentation are tightly coupled. Through an iterative optimization process, more and more correct correspondences are detected owing to object boundaries revealed by co-segmentation. The proposed approach is comprehensively evaluated. Promising experimental results on four datasets manifest its effectiveness.
Due to the difficulty of creating pitch-labeled training data that cover the rich diversity found in music signals, unsupervised feature-based approaches derived from signal processing and feature design remain critical for multipitch estimation (MPE) of polyphonic music. While a large number of feature representations have been proposed in the literature, an effective means of combining different domains of features for MPE is still needed. In this paper, we propose a novel approach, referred to as combined frequency and periodicity (CFP), that detects pitches according to the agreement of a harmonic series in the frequency domain and a subharmonic series in the lag (quefrency) domain. This approach nicely aggregates the complementary advantages of the two feature domains in different frequency ranges, and improves the robustness of the pitch detection function to the interference of the overtones of simultaneous pitches. We report a comprehensive evaluation that compares CFP against three state-of-the-art approaches using three MPE datasets and four symphonies. The evaluation is characteristic of the coverage and complexity of music (in terms of instrument type and degree of polyphony). In addition, we also evaluate the performance of the MPE approaches when a number of audio degradations are applied. Results show that the proposed unsupervised method performs consistently well across the types of Western polyphonic music considered, and is robust to audio degradations such as high-pass filtering and MP3 compression.
N/A
Mobile devices have increasingly been used to run multimedia applications which are extremely downlink-intensive. The conventional rate adaptive and/or margin adaptive approach for radio resource allocation may result in unnecessary energy consumption on mobile devices, which will not be energy efficient for mobile multimedia applications. In this paper, we develop an energy adaptive approach and design an energy-efficient downlink resource allocation scheme to support multimedia applications. The objective is to minimize the total energy consumption of mobile devices for data reception while meeting the data rate requirements at mobile devices and the transmit power constraint at the base station. We show that the optimization problem is NP -hard and then propose an efficient algorithm that has a provable performance guarantee under a certain condition. We have conducted extensive simulations to evaluate the efficacy of the proposed algorithm and our results provide useful insights into the design of energy-efficient resource allocation for wireless systems.
none
Unsupervised domain adaptation (UDA) deals with the task that labeled training and unlabeled test data col- lected from source and target domains, respectively. In this paper, we particularly address the practical and challenging scenario of imbalanced cross-domain data. That is, we do not assume the label numbers across do- mains to be the same, and we also allow the data in each domain to be collected from multiple datasets/sub- domains. To solve the above task of imbalanced domain adaptation , we propose a novel algorithm of Domain- constraint Transfer Coding (DcTC) . Our DcTC is able to exploit latent subdomains within and across data do- mains, and learns a common feature space for joint adaptation and classification purposes. Without assum- ing balanced cross-domain data as most existing UDA approaches do, we show that our method performs fa- vorably against state-of-the-art methods on multiple cross-domain visual classification tasks.
This paper studies joint subchannel allocation, power allocation, and beamforming for simultaneous wireless information and power transfer (SWIPT) in multiuser downlink orthogonal frequency-division multiple access (OFDMA) systems. We formulate a multi-objective optimization (MOO) problem where the objectives are to maximize both the information rate and the harvested power for all users in the system. We approach the MOO problem with two proposed methods, i.e., semidefinite relaxation based weighted aggregation (SDR-WA) and multi-objective genetic algorithm (MOGA). Simulation compares the achievable Pareto optimal solution set yielded by these methods, and illustrates the tradeoffs of the sum information rate vs. the sum harvested power in the system.
Graphics-intensive mobile games are becoming increasingly popular, but such applications place high demand on device CPUs and GPUs. The design of current mobile systems results in unnecessary energy waste due to lack of consideration of application phases and user attention (a “demand-level” gap) and because each processor administers power management autonomously (a “processor-level” gap). This paper proposes a user-centric CPU-GPU governing framework which aims to reduce energy consumption without significantly impacting the user experience. To bridge the gap at the demand level, we identify the user demand at runtime and accordingly determine appropriate governing policies for the respective processors. On the other hand, to bridge the gap at the processor level, the proposed framework interprets the frequency scaling intents of processors based on the observation of the CPU-GPU interaction and the processor status. We implemented our framework on a Samsung Galaxy S4, and conducted extensive experiments with real-world 3D gaming apps. Experimental results showed that, for an application being highly interactive and frequent phase changing, our proposed framework can reduce energy consumption by 45.1% compared with state-of-the-art policy without significantly impacting the user experience.
N/A
This paper studies the information-theoretic secrecy rates of wireless two-way relay systems where two users wish to exchange information through a single relay with an eavesdropper observing all communications. We formulate and compare the achievable secrecy rates of the system that employs one of the three common relay protocols: conventional decode-and-forward (DF), DF with network coding (NC), and compute-and-forward (CF) based on physical-layer network coding (PNC). We show that CF based on PNC achieves the highest secrecy rate at high signal-to-noise ratio (SNR), while, interestingly, the other two protocols have mixed performance depending on the power allocation scheme and the network topology. Our study offers insights into designing wireless two-way relay protocols from a secrecy perspective.