Most recent semi-supervised video object segmentation (VOS) methods rely on fine-tuning deep convolutional neural networks online using the given mask of the first frame or predicted masks of subsequent frames. However, the online fine-tuning process is usually time-consuming, limiting the practical use of such methods. We propose a directional deep embedding and appearance learning (DDEAL) method, which is free of the online fine-tuning process, for fast VOS. First, a global directional matching module, which can be efficiently implemented by parallel convolutional operations, is proposed to learn a semantic pixel-wise embedding as an internal guidance. Second, an effective directional appearance model based statistics is proposed to represent the target and background on a spherical embedding space for VOS. Equipped with the global directional matching module and the directional appearance model learning module, DDEAL learns static cues from the labeled first frame and dynamically updates cues of the subsequent frames for object segmentation. Our method exhibits state-of-the-art VOS performance without using online fine-tuning. Specifically, it achieves a & mean score of 74.8% on DAVIS 2017 dataset and an overall score of 71.3% on the large-scale YouTube-VOS dataset, while retaining a speed of 25 fps with a single NVIDIA TITAN Xp GPU. Furthermore, our faster version runs 31 fps with only a little accuracy loss.

Ieee transactions on neural networks and learning systems

Before the emergence of ultra-wideband (UWB) radios, widely used wireless communications were based on sinusoidal carriers, and impulse technologies were employed only in specific applications (e.g. radar). In 2002, the Federal Communication Commission (FCC) allowed unlicensed operation between 3.1–10.6 GHz for UWB communication, using a wideband signal format with a low EIRP level (−41.3dBm/MHz). UWB communication systems then emerged as an alternative to narrowband systems and significant effort in this area has been invested at the regulatory, commercial, and research levels.

IEEE Transactions on Microwave Theory and Techniques and Antennas and Propagation Announce a Joint Special Issue on Ultra-Wideband (UWB) Technology

The continuously increasing demand for wide bandwidth creates great difficulties in employing digital predistortion (DPD) for radio frequency (RF) power amplifiers (PAs) in future ultra-wideband systems because the existing DPD system requires multiple times the input signal bandwidth in the transmitter and receiver chain, which is sometimes almost impossible to implement in practice. In this paper, we present a novel band-limited digital predistortion technique in which a band-limiting function is inserted into the general Volterra operators in the DPD model to control the signal bandwidth under modeling, which logically transforms the general Volterra series-based model into a band-limited version. This new approach eliminates the system bandwidth constraints of the conventional DPD techniques, and it allows users to arbitrarily choose the bandwidth to be linearized in the PA output according to the system requirement without sacrificing performance, which makes the DPD system design much more flexible and feasible. In order to validate this idea, a high-power LDMOS Doherty PA excited by various wideband signals, including 100-MHz long-term evolution advanced signals, was tested. Experimental results showed that excellent linearization performance can be obtained by employing the proposed approach. Furthermore, this technique can be applied to other linear-in-parameter models. In future ultra-wideband systems, this new technique can significantly improve system performance and reduce DPD implementation cost.

/pdf/band-limited-volterra-series-based-digital-predistortion-for-43umodaj9s.pdf

Band-Limited Volterra Series-Based Digital Predistortion for Wideband RF Power Amplifiers

The RF PA, as one of the most essential components in any wireless system, suffers from inherent nonlinearities. The output of a PA must comply with the linearity requirement specified by the standards. Due to its satisfactory linearization capability, DPD has been widely accepted as one of the fundamental units in modern and future wideband wireless systems. With the help of this flexible digital technology, the inherent linearity problem of PAs operating in the saturation region can be significantly improved, which enables us, the wireless engineers, to create more suitable RF transceiver architectures to provide wireless access with better user experience (linearity perspective) and less power waste (power efficiency perspective). This moves us one more step towards the ultimate green communications. In this article, we discussed the DPD techniques in the context of linearizing nonlinear RF PAs. As the computing-horsepower and the transistor-density of digital IC increases while the cost per transistor decreases, the concept that uses digital enhancement techniques to eliminate active analog imperfects will gain more attention from both industry and academia.

/pdf/green-communications-digital-predistortion-for-wideband-rf-1soybypt4y.pdf

Green Communications: Digital Predistortion for Wideband RF Power Amplifiers

A new behavioral model for digital predistortion of radio frequency (RF) power amplifiers (PAs) is proposed in this paper. It is derived from a modified form of the canonical piecewise-linear (CPWL) functions using a decomposed vector rotation (DVR) technique. In this model, the nonlinear basis function is constructed from piecewise vector decomposition, which is completely different from that used in the conventional Volterra series. Theoretical analysis has shown that this model is much more flexible in modeling RF PAs with non-Volterra-like behavior, and experimental results confirmed that the new model can produce excellent performance with a relatively small number of coefficients when compared to conventional models.

https://researchrepository.ucd.ie/bitstream/10197/8424/1/TMTT-2014-09-0998_Final.pdf

Decomposed Vector Rotation-Based Behavioral Modeling for Digital Predistortion of RF Power Amplifiers

A simple and flexible technique for improving the modeling and predistortion of power amplifiers is presented. The technique, which relies on the sparsity assumption for the kernel coefficients of the full Volterra (FV) behavioral model, combines a greedy algorithm for the selection of the active coefficients, a maximum likelihood method for their estimation and an information criterion for determining the best model. The approach has been applied to the design of reduced-parameters FV-based digital predistorters for three power amplifiers driven with orthogonal frequency division multiplexing signals, following the LTE and DVB-T2 standards, and a multichannel wideband code-division multiple access signal. Results show that the proposed linearizers meet the spectral masks and error vector magnitude constraints of the referred standards and provide a reduction better than 45% in the number of parameters, compared to the FV predistorters.

Behavioral Modeling and Predistortion of Power Amplifiers Under Sparsity Hypothesis

A single envelope modulator-based envelope tracking (ET) power amplifier (PA) structure with a related digital predistortion (DPD) technique for multiple-input and multiple-output (MIMO) wireless transmitters is presented in this paper. By generating a common tracking envelope, only one envelope modulator is employed for controlling supply voltage of the RF PAs in all branches in the system, which dramatically reduces the system implementation cost. Due to the structure change, additional distortion is introduced, and it is difficult to directly compensate by using the conventional DPD because the tracking envelope is no longer the same as the RF envelope, and thus the MIMO ET PA becomes a two-input and one-output system. To resolve this problem, in this paper we propose a novel DPD technique in which the PA input and output data are reconstructed into multiple data subsets according to variations of the tracking envelope. It converts the 2-to-1 mapping into multiple 1-to-1 ones, where the conventional DPD can be employed again. Experimental results demonstrated that the distortion, including static nonlinearities, memory effects, and additional distortion caused by the structure change can be effectively compensated by using the proposed DPD technique. Compared to the conventional ET, the overall efficiency of the system is only slightly decreased, but the system cost is much lower because only one envelope modulator is required in the whole system.

https://researchrepository.ucd.ie/bitstream/10197/8644/1/TMTT201210_ChaoYu_Final_Draft.pdf

A Single Envelope Modulator-Based Envelope-Tracking Structure for Multiple-Input and Multiple-Output Wireless Transmitters

In this paper, a bandwidth-constrained least squares (LS)-based model extraction is proposed to enhance the linearization performance of band-limited digital predistortion (DPD) for RF power amplifiers (PAs). By laying frequency bandwidth constraints on the conventional LS-based model extraction, the modified extraction method is adapted for the band-limited DPD scenario. Compared to the conventional LS-based model extraction, the proposed bandwidth-constrained LS based approach can provide an accurate model extraction with much narrower bandwidth. In order to evaluate the performance, experimental validation was carried out on a wideband Doherty PA, which was designed for 3G/4G wireless communication systems. By applying the proposed approach, the ACLR achieved below −55 dBc for 40 MHz 8-carrier UMTS signal within only 80 MHz observation bandwidth.

Bandwidth-constrained least squares-based model extraction for band-limited digital predistortion of RF power amplifiers

In this paper, a novel single-model single-feedback digital predistortion (DPD) technique is proposed to linearize concurrent multi-band wireless transmitters. By employing carrier relocation and band-limiting functions in digital signal processing, DPD model complexity is significantly reduced and only one narrowband feedback loop is required for data acquisition. In the new system, the linearization bandwidth can be arbitrarily reconfigured, which provides considerable flexibility for DPD system design in wideband concurrent operations. Experimental tests were conducted to validate various cases including a concurrent quad-band scenario that is reported for the first time to date. Excellent performance demonstrates that the proposed approach provides an effective solution for reducing DPD system implementation complexity and cost in both digital and analog domains for future wideband concurrent multi-band transmitters.

https://researchrepository.ucd.ie/bitstream/10197/8401/1/TMTT-2014-10-1227_Final.pdf

Chao Yu

Papers

Band-Limited Volterra Series-Based Digital Predistortion for Wideband RF Power Amplifiers

Behavioral Modeling and Predistortion of Power Amplifiers Under Sparsity Hypothesis

A Single Envelope Modulator-Based Envelope-Tracking Structure for Multiple-Input and Multiple-Output Wireless Transmitters

Bandwidth-constrained least squares-based model extraction for band-limited digital predistortion of RF power amplifiers

Single-Model Single-Feedback Digital Predistortion for Concurrent Multi-Band Wireless Transmitters