Shimin Zhang

Xianwei Chen, Shimin Zhang, Jibin Wu

Scaling on-policy distillation (OPD) for large language models (LLMs) confronts a fundamental tension: asynchronous execution is necessary for system efficiency, but structurally deviates from the ideal on-policy objective. To address this challenge, we theoretically decompose the objective discrepancy into rollout drift and supervision drift, capturing staleness in student rollout and teacher context, respectively. Building on this, we introduce a sample-level freshness score that quantifies the reliability of a buffered sample with respect to the on-policy objective. Guided by this signal, we further propose f-OPD, a novel framework that adaptively regulates stale-sample influence and constrains policy drift accumulated under asynchronous training. Across reasoning, tool-use, and coding-agent tasks of increasing interaction horizon, f-OPD consistently achieves task performance comparable to synchronous optimization while largely retaining the throughput advantages of asynchronous execution. Our results establish the first recipe for achieving a performance-efficiency trade-off in OPD, paving the way for long-horizon agentic post-training at scale.

Shimin Zhang, Yechuan Xia, Chunxiao Li et al.

Over the past several decades, CDCL SAT solvers have proven remarkably effective on large industrial formulas, despite SAT being NP-complete and widely believed to be intractable. While considerable empirical research has been done on solver performance over benchmarks like the SAT competition, as well as scaling studies on random and crafted families, surprisingly little effort has gone into systematic scaling studies over industrial instances. To address this gap, we collect a large benchmark of Bounded Model Checking (BMC) instances (76,600+ across 766 families) and perform a systematic scaling study of solver performance. We observe a spectrum: some families scale linearly, others polynomially or exponentially. Building on this foundation, we study the structural parameters that have been proposed to explain this phenomenon. We first show that previously proposed parameters -- clause-variable ratio, treewidth, and community structure -- fail to discriminate between the linear and exponential regimes. By contrast, the recently proposed \emph{proofdoor} parameter explains this phenomenon well. Informally, a proofdoor is a sequence of interpolants between chunks of a formula, where each interpolant represents the solver's memoization of reasoning effort on chunks it has already analyzed. In support of the proofdoor hypothesis, we make three key contributions. First, we empirically show that CDCL solvers do compute small proofdoors for linearly-scaling BMC instances. Second, we show that for exponentially-scaling instances, sampled proofdoors scale exponentially and are typically not incrementally absorbed. Third, we show that scrambling linearly-scaling instances yields larger proofdoor sizes relative to pre-scrambling, relating poor branching order to larger proofdoor sizes and drop in solver performance.

Zeyang Song, Shimin Zhang, Yuhong Chou et al.

Spiking Neural Networks (SNNs), inspired by biological neural mechanisms, represent a promising neuromorphic computing paradigm that offers energy-efficient alternatives to traditional Artificial Neural Networks (ANNs). Despite proven effectiveness, SNN architectures have struggled to achieve competitive performance on large-scale speech processing tasks. Two key challenges hinder progress: (1) the high computational overhead during training caused by multi-timestep spike firing, and (2) the absence of large-scale SNN architectures tailored to speech processing tasks. To overcome the issues, we introduce Input-aware Multi-Level Spikeformer, i.e. IML-Spikeformer, a spiking Transformer architecture specifically designed for large-scale speech processing. Central to our design is the Input-aware Multi-Level Spike (IMLS) mechanism, which simulates multi-timestep spike firing within a single timestep using an adaptive, input-aware thresholding scheme. IML-Spikeformer further integrates a Re-parameterized Spiking Self-Attention (RepSSA) module with a Hierarchical Decay Mask (HDM), forming the HD-RepSSA module. This module enhances the precision of attention maps and enables modeling of multi-scale temporal dependencies in speech signals. Experiments demonstrate that IML-Spikeformer achieves word error rates of 6.0\% on AiShell-1 and 3.4\% on Librispeech-960, comparable to conventional ANN transformers while reducing theoretical inference energy consumption by 4.64$\times$ and 4.32$\times$ respectively. IML-Spikeformer marks an advance of scalable SNN architectures for large-scale speech processing in both task performance and energy efficiency. Our source code and model checkpoints are publicly available at github.com/Pooookeman/IML-Spikeformer.

Yihui Fu, Zhuoyuan Yao, Weipeng He et al.

The IEEE Spoken Language Technology Workshop (SLT) 2021 Alpha-mini Speech Challenge (ASC) is intended to improve research on keyword spotting (KWS) and sound source location (SSL) on humanoid robots. Many publications report significant improvements in deep learning based KWS and SSL on open source datasets in recent years. For deep learning model training, it is necessary to expand the data coverage to improve the robustness of model. Thus, simulating multi-channel noisy and reverberant data from single-channel speech, noise, echo and room impulsive response (RIR) is widely adopted. However, this approach may generate mismatch between simulated data and recorded data in real application scenarios, especially echo data. In this challenge, we open source a sizable speech, keyword, echo and noise corpus for promoting data-driven methods, particularly deep-learning approaches on KWS and SSL. We also choose Alpha-mini, a humanoid robot produced by UBTECH equipped with a built-in four-microphone array on its head, to record development and evaluation sets under the actual Alpha-mini robot application scenario, including noise as well as echo and mechanical noise generated by the robot itself for model evaluation. Furthermore, we illustrate the rules, evaluation methods and baselines for researchers to quickly assess their achievements and optimize their models.

Shimin Zhang, Xianwei Chen, Yufan Shen et al.

Reinforcement Learning with Verifiable Rewards (RLVR) has recently demonstrated remarkable potential in enhancing the reasoning capability of Large Reasoning Models (LRMs). However, RLVR often leads to entropy collapse, resulting in premature policy convergence and performance saturation. While manipulating token-level entropy has proven effective for promoting policy exploration, we argue that the latent dynamics underlying token generation encode a far richer computational structure for steering policy optimization toward a more effective exploration-exploitation tradeoff. To enable tractable analysis and intervention of the latent dynamics of LRMs, we leverage Koopman operator theory to obtain a linearized representation of their hidden-state dynamics. This enables us to introduce Dynamic Spectral Dispersion (DSD), a new metric to quantify the heterogeneity of the model's latent dynamics, serving as a direct indicator of policy exploration. Building upon these foundations, we propose Reasoning with Latent eXploration (ReLaX), a paradigm that explicitly incorporates latent dynamics to regulate exploration and exploitation during policy optimization. Comprehensive experiments across a wide range of multimodal and text-only reasoning benchmarks show that ReLaX significantly mitigates premature convergence and consistently achieves state-of-the-art performance.

Shimin Zhang, Ziteng Wang, Jiayao Sun et al.

This paper introduces the NWPU Team's entry to the ICASSP 2022 AEC Challenge. We take a hybrid approach that cascades a linear AEC with a neural post-filter. The former is used to deal with the linear echo components while the latter suppresses the residual non-linear echo components. We use gated convolutional F-T-LSTM neural network (GFTNN) as the backbone and shape the post-filter by a multi-task learning (MTL) framework, where a voice activity detection (VAD) module is adopted as an auxiliary task along with echo suppression, with the aim to avoid over suppression that may cause speech distortion. Moreover, we adopt an echo-aware loss function, where the mean square error (MSE) loss can be optimized particularly for every time-frequency bin (TF-bin) according to the signal-to-echo ratio (SER), leading to further suppression on the echo. Extensive ablation study shows that the time delay estimation (TDE) module in neural post-filter leads to better perceptual quality, and an adaptive filter with better convergence will bring consistent performance gain for the post-filter. Besides, we find that using the linear echo as the input of our neural post-filter is a better choice than using the reference signal directly. In the ICASSP 2022 AEC-Challenge, our approach has ranked the 1st place on word accuracy (WAcc) (0.817) and the 3rd place on both mean opinion score (MOS) (4.502) and the final score (0.864).

Yusheng Wang, Weiwei Song, Yidong Lou et al.

Precise and real-time rail vehicle localization as well as railway environment monitoring is crucial for railroad safety. In this letter, we propose a multi-LiDAR based simultaneous localization and mapping (SLAM) system for railway applications. Our approach starts with measurements preprocessing to denoise and synchronize multiple LiDAR inputs. Different frame-to-frame registration methods are used according to the LiDAR placement. In addition, we leverage the plane constraints from extracted rail tracks to improve the system accuracy. The local map is further aligned with global map utilizing absolute position measurements. Considering the unavoidable metal abrasion and screw loosening, online extrinsic refinement is awakened for long-during operation. The proposed method is extensively verified on datasets gathered over 3000 km. The results demonstrate that the proposed system achieves accurate and robust localization together with effective mapping for large-scale environments. Our system has already been applied to a freight traffic railroad for monitoring tasks.

Yusheng Wang, Yidong Lou, Yi Zhang et al.

We present RailLoMer in this article, to achieve real-time accurate and robust odometry and mapping for rail vehicles. RailLoMer receives measurements from two LiDARs, an IMU, train odometer, and a global navigation satellite system (GNSS) receiver. As frontend, the estimated motion from IMU/odometer preintegration de-skews the denoised point clouds and produces initial guess for frame-to-frame LiDAR odometry. As backend, a sliding window based factor graph is formulated to jointly optimize multi-modal information. In addition, we leverage the plane constraints from extracted rail tracks and the structure appearance descriptor to further improve the system robustness against repetitive structures. To ensure a globally-consistent and less blurry mapping result, we develop a two-stage mapping method that first performs scan-to-map in local scale, then utilizes the GNSS information to register the submaps. The proposed method is extensively evaluated on datasets gathered for a long time range over numerous scales and scenarios, and show that RailLoMer delivers decimeter-grade localization accuracy even in large or degenerated environments. We also integrate RailLoMer into an interactive train state and railway monitoring system prototype design, which has already been deployed to an experimental freight traffic railroad.

Shubo Lv, Yanxin Hu, Shimin Zhang et al.

Deep complex convolution recurrent network (DCCRN), which extends CRN with complex structure, has achieved superior performance in MOS evaluation in Interspeech 2020 deep noise suppression challenge (DNS2020). This paper further extends DCCRN with the following significant revisions. We first extend the model to sub-band processing where the bands are split and merged by learnable neural network filters instead of engineered FIR filters, leading to a faster noise suppressor trained in an end-to-end manner. Then the LSTM is further substituted with a complex TF-LSTM to better model temporal dependencies along both time and frequency axes. Moreover, instead of simply concatenating the output of each encoder layer to the input of the corresponding decoder layer, we use convolution blocks to first aggregate essential information from the encoder output before feeding it to the decoder layers. We specifically formulate the decoder with an extra a priori SNR estimation module to maintain good speech quality while removing noise. Finally a post-processing module is adopted to further suppress the unnatural residual noise. The new model, named DCCRN+, has surpassed the original DCCRN as well as several competitive models in terms of PESQ and DNSMOS, and has achieved superior performance in the new Interspeech 2021 DNS challenge

Shimin Zhang, Yuxiang Kong, Shubo Lv et al.

With the increasing demand for audio communication and online conference, ensuring the robustness of Acoustic Echo Cancellation (AEC) under the complicated acoustic scenario including noise, reverberation and nonlinear distortion has become a top issue. Although there have been some traditional methods that consider nonlinear distortion, they are still inefficient for echo suppression and the performance will be attenuated when noise is present. In this paper, we present a real-time AEC approach using complex neural network to better modeling the important phase information and frequency-time-LSTMs (F-T-LSTM), which scan both frequency and time axis, for better temporal modeling. Moreover, we utilize modified SI-SNR as cost function to make the model to have better echo cancellation and noise suppression (NS) performance. With only 1.4M parameters, the proposed approach outperforms the AEC-challenge baseline by 0.27 in terms of Mean Opinion Score (MOS).

Yanxin Hu, Yun Liu, Shubo Lv et al.

Speech enhancement has benefited from the success of deep learning in terms of intelligibility and perceptual quality. Conventional time-frequency (TF) domain methods focus on predicting TF-masks or speech spectrum, via a naive convolution neural network (CNN) or recurrent neural network (RNN). Some recent studies use complex-valued spectrogram as a training target but train in a real-valued network, predicting the magnitude and phase component or real and imaginary part, respectively. Particularly, convolution recurrent network (CRN) integrates a convolutional encoder-decoder (CED) structure and long short-term memory (LSTM), which has been proven to be helpful for complex targets. In order to train the complex target more effectively, in this paper, we design a new network structure simulating the complex-valued operation, called Deep Complex Convolution Recurrent Network (DCCRN), where both CNN and RNN structures can handle complex-valued operation. The proposed DCCRN models are very competitive over other previous networks, either on objective or subjective metric. With only 3.7M parameters, our DCCRN models submitted to the Interspeech 2020 Deep Noise Suppression (DNS) challenge ranked first for the real-time-track and second for the non-real-time track in terms of Mean Opinion Score (MOS).