The prevailing Gaussian Splatting has been adapted for image representation and compression by GaussianImage recently, achieving impressive rendering speeds of 1500–2000 FPS. However, it falls short in rate-distortion performance compared to current state-of-the-art image codecs. We attribute this to its explicit representation and tile-based rasterization process which demands a more efficient utilization of 2D Gaussians for improved performance. In this work, we propose a spectrum-adaptive distribution method to allocate Gaussians in alignment with the complexity of varying image regions. Additionally, we introduce a gradient-based growing strategy for Gaussians to further refine their distribution. Experimental results show that our approach outperforms GaussianImage in image representation quality while maintaining comparable training and rendering speeds. Moreover, by integrating a transform-quantization pipeline for the position attributes of Gaussians, our approach delivers a better rate-distortion curve compared to the INR-based codecs such as COIN and COIN++.

Existing solutions to the joint problem of image deblurring and compression typically adopt a straightforward sequential pipeline, where the two task are handled independently by either compressing the blurred image first or by deblurring before compression. However, such decoupled strategies often suffer from error accumulation and increased computational overhead, leading to suboptimal performance in both compression and restoration. In this work, we propose a joint solution to simultaneously address image deblurring and compression within a learned image compression framework, with the capacity to encode a compact bitstream of the underlying sharp image from its blurred observation. Specifically, our framework has a two-branch encoder architecture, consisting of a main branch and a guidance branch. The guidance branch extracts blur-aware features from the blurry image, which are then fused into the main branch to assist in encoding features that are both bitrate-efficient and essential for accurate sharp restoration. Furthermore, we enhance the fused features in both the spatial and frequency domains. In particular, a JPEG-like transform-quantization mechanism is adopted to modulate the fused features using differentiable quantization matrices in the frequency domain. Extensive experiments on widely-used deblurring datasets demonstrate that our joint solution achieves superior rate-distortion performance, while significantly reducing inference latency compared to existing sequential approaches.

The generalized zero-shot diagnosis (GZSD) has attracted rising attention from researchers, but faces a major bottleneck: the domain shift problem (DSP), which causes most unseen faults to be misclassified as seen. In this study, we propose a novel method termed semisupervised hybrid triplet network (SSHTN) with high resistance toward DSP to achieve GZSD by first introducing auxiliary set. To leverage the unlabeled data of auxiliary set, which is different from known faults (both seen and unseen), SSHTN defines relative boundaries of seen and unseen faults in the similarity space and comprises two branches: the data–data branch (DDB) and the data–semantic branch (DSB). For DDB, it constructs homogeneous and heterogeneous data pairs to learn data–data similarity. With the support of auxiliary set, the semisupervised similarity learning mechanism is constructed. Specifically, DDB analyzes differences between data pairs mixed by seen and auxiliary fault data to narrow down the decision boundaries of seen faults. Meanwhile, similarities between data pairs formed by disturbing auxiliary set data are learned to improve the generalization of DDB. For DSB, it learns data–semantic similarity by matching fault data with corresponding descriptions and extends SSHTN discriminative ability toward unseen faults. The SSHTN obtained from the comprehensive learning of similarity by the two branches could determine the boundaries of seen and unseen faults more accurately, thereby alleviating the effect of DSP. Experiments are implemented on the Tennessee Eastman process and a real thermal power plant process, and results show the superior performance of SSHTN over state-of-the-art ZSD methods.

The time delay of digital control would make the equivalent output admittance of voltage source converter(VSC) non-passive, which may lead to the resonant instable problem. In this paper, the Luenberger observer-based active damping method is researched to make the admittance passive. The method contains two part: firstly, the observer predicts the capacitor current of the LCL filter; secondly, carrying out the active damping by using the predicted current. It turns out that the passivity of VSC's equivalent admittance is affected by both the observer and active damping, and the passive equivalent admittance can be obtained by properly designing the observer and active damping parameters. In practical application, the observer-based active damping method can not only avoid using additional current sensors but also make the equivalent admittance passive below the Nyquist frequency. The simulation results verified the effectiveness of the proposed method for improving stability.

This paper proposes an interpolation based progressive algebraic Chase decoding (PACD) algorithm for Reed-Solomon (RS) codes. Based on the received information, 2η (η > 0) interpolation test-vectors are constructed. They are ordered using a reliability function, assessing their potential of yielding the intended message. The decoding is performed progressively granting priority to decode the test-vectors that are more likely to yield the intended message, and it will be terminated once the intended message is found. In the proposal, the decoding of a later test-vector utilizes the interpolation information that is generated during the decoding of the earlier ones. It results in the binary tree that represents the evolution of the interpolated polynomial sets growing in a depth-first-search manner. The PACD algorithm has the advantage of adapting its decoding computation to the channel condition, leveraging the average decoding complexity. This channel dependent feature will be validated by our simulation results which show that the PACD algorithm is less complex than various interpolation based algebraic decoding algorithms. We will also demonstrate that it can achieve a high RS decoding performance.

This study proposes a progressive algebraic Chase decoding (PACD) algorithm for Reed–Solomon (RS) codes. On the basis of the received information, 2 η ( η > 0) interpolation test‐vectors are constructed for the interpolation‐based algebraic Chase decoding. A test‐vector reliability function is defined to assess their potential for yielding the intended message. The algebraic Chase decoding will then be performed progressively granting priority to decode the test‐vectors that are more likely to yield the message, and is then terminated once it is found. Consequently, the decoding complexity can be adapted to the quality of the received information. An enhanced‐PACD (E‐PACD) algorithm is further proposed by coupling the PACD algorithm with the adaptive belief propagation (ABP) decoding. The ABP decoding generates new test‐vectors for the PACD algorithm by enhancing the received information. It improves the Chase decoding performance without increasing the decoding complexity exponentially. It is shown that the E‐PACD algorithm's complexity can be significantly reduced by utilising the existing interpolation information of the previous Chase decodings’. Our performance evaluations show that the two proposed decoders outperform a number of existing algebraic decoding approaches. Complexity and memory analyses of the PACD algorithm are also presented, demonstrating that this is an efficient RS decoding strategy.