The emergence of multimodal technologies has propelled Vision-Language Incremental Learning (VLIL) into a research spotlight. Current VLIL approaches predominantly inherit unimodal paradigms, failing to address fundamental distinctions between visual and linguistic modalities. Crucially, the semantic gap between images and text creates divergent learning dynamics: visual data exhibits rich, distributed information while textual representations remain explicit and compact. Consequently, textual elements align with class-specific tasks, whereas individual images inherently span multiple such tasks, creating dual bottlenecks in class-level memory allocation and scene-level knowledge transfer. To overcome these challenges, we propose DCIM (Dual Class-Individual Memory), a novel framework featuring complementary mechanisms for vision-language continual learning. For class-level constraints, our Hierarchical Class Memory Management (HCMM) strategy dynamically allocates memory resources across object categories. It employs forgetting simulation to identify and preserve the most vulnerable samples, ensuring robust long-term knowledge retention. For scene-level adaptation, the Scene Reconstruction Memory(SRM) module captures generalized environmental representations, enabling contextual transfer to novel classes and disambiguation of semantically related concepts within shared scenes.Extensive experiments on two vision-language tasks, i.e., visual question answering (VQA) and Image captioning (IC), demonstrate the effectiveness and excellent generalization ability of our approach, achieving superior performance under continual learning settings.

With the rise of virtual communication and smart devices, speech has become the most natural medium of interaction. Yet it remains intrinsically difficult: speech is fleeting, unstructured, and disfluent, making key information prone to loss. Conventional Speech-to-Text (STT) systems attempt to acoustically reconstruct what was said. However, their frame-level alignment and rigid token-by-token decoding break down under noise, interruptions, or fragmentation. Humans, in contrast, readily grasp what was meant by exploiting syntax, discourse, pragmatics, and prosody. We argue for a paradigm shift from acoustic reconstruction to semantic transduction: inferring meaning directly from speech, abstracted from surface distortions. This shift raises two challenges: (C1) the lack of anchors between audio and meaning, and (C2) the need to maintain compositional semantics. To address these, we introduce CogTrans, a cognitively inspired speech-to-meaning framework. CogTrans tackles C1 through a Semantic Anchor Explorer, built on I-JEPA to capture higher-order regularities, prosodic rhythms, cross-frequency coarticulation, discourse continuity-providing resilient semantic scaffolds under noise and fragmentation. For C2, it designs a Lexical-Semantic Harmonizer that dynamically integrates these anchors with lexical embeddings; thereby preserving fine-grained compositional fidelity in roles, order, and entities. Extensive experiments show that CogTrans delivers consistent and substantial gains under challenging conditions. On GigaSpeech, it achieves a 6.58% relative Word Error Rate (WER) reduction, and on the multilingual VoxPopuli benchmark, the gain climbs to 12.97% at 10 dB noise-a regime where conventional models typically collapse. Beyond literal accuracy, CogTrans also boosts semantic fidelity, with a 3.40% increase in ROUGE-L and 3.45% in USE-Sim, ensuring transcripts remain faithful not only in words but also in meaning. Together, these results underscore that CogTrans is robust in noisy, unconstrained environments-precisely the conditions where reliability matters most.

*Multi-agent reinforcement learning* (MARL) has been shown effective for cooperative games in recent years. However, existing state-of-the-art methods face challenges related to sample complexity, training instability, and the risk of converging to a suboptimal Nash Equilibrium. In this paper, we propose a unified framework for learning \emph{stochastic} policies to resolve these issues. We embed cooperative MARL problems into probabilistic graphical models, from which we derive the maximum entropy (MaxEnt) objective for MARL. Based on the MaxEnt framework, we propose *Heterogeneous-Agent Soft Actor-Critic* (HASAC) algorithm. Theoretically, we prove the monotonic improvement and convergence to *quantal response equilibrium* (QRE) properties of HASAC. Furthermore, we generalize a unified template for MaxEnt algorithmic design named *Maximum Entropy Heterogeneous-Agent Mirror Learning* (MEHAML), which provides any induced method with the same guarantees as HASAC. We evaluate HASAC on six benchmarks: Bi-DexHands, Multi-Agent MuJoCo, StarCraft Multi-Agent Challenge, Google Research Football, Multi-Agent Particle Environment, and Light Aircraft Game. Results show that HASAC consistently outperforms strong baselines, exhibiting better sample efficiency, robustness, and sufficient exploration.

Reconfigurable intelligent Surface (RIS) are often proposed as a means for restoring non-line-of-sight (NLoS) links in wireless communication. In this paper, we propose a joint design strategy for multi-RIS deployment locations and passive beamforming in multi-RIS assisted multi-BS networks with blockages. To ensure a stable network topology and low feedback overhead, the ergodic capacity is analyzed using Jensen’s inequality and the binomial expansion theorem, and then maximized as a performance metric. Taking into account the minimum spacing constraint for multi-RIS in real-world scenarios, we optimize the multi-RIS deployment locations using a genetic algorithm, where a Markov chain model can be applied to obtain the global optimal solution. Additionally, we implement the passive beamforming design for multi-RIS using the Riemannian conjugate gradient (RCG) method based on manifold space. Numerical results demonstrate the superiority of our proposed deployment strategy over the benchmark deployment schemes.

As an increasingly popular flow metering technology, Coriolis mass flowmeter exhibits high measurement accuracy under single-phase flow condition and is widely used in the industry. However, under complex flow conditions, such as two-phase flow, the measurement accuracy is greatly decreased due to various factors including improper signal processing methods. In this study, three digital signal processing methods—the quadrature demodulation (QD) method, Hilbert method, and sliding discrete time Fourier transform method—are analyzed for their applications in processing sensor signals and providing measurement results under gas-liquid two-phase flow condition. Based on the analysis, specific improvements are applied to each method to deal with the signals under two-phase flow condition. For simulation, sensor signals under single- and two-phase flow conditions are established using a random walk model. The phase difference tracking performances of these three methods are evaluated in the simulation. Based on the digital signal processor, a converter program is implemented on its evaluation board. The converter program is tested under single- and two-phase flow conditions. The improved signal processing methods are evaluated in terms of the measurement accuracy and complexity. The QD algorithm has the best performance under the single-phase flow condition. Under the two-phase flow condition, the QD algorithm performs a little better in terms of the indication error and repeatability than the improved Hilbert algorithm at 160, 250, and 420 kg/h flow points, whereas the Hilbert algorithm outperforms the QD algorithm at the 600 kg/h flow point.