BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Nuo Chen

Nuo Chen contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
Computation and LanguageArtificial Intelligenceeess.ASSoundComputer Visioneess.SPInformation Theorymath.ITastro-ph.GAMultimediaphysics.optics

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
13works
0followers
11topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

13 published item(s)

preprint2026arXiv

EmoMM: Benchmarking and Steering MLLM for Multimodal Emotion Recognition under Conflict and Missingness

Multimodal Emotion Recognition (MER) is critical for interpreting real-world interactions. While Multimodal Large Language Models (MLLM) have shown promise in MER, their internal decision-making mechanisms under modality conflict and missingness remain largely underexplored. In this paper, to systematically investigate these behaviors, we introduce EmoMM, a comprehensive benchmark featuring modality-aligned, conflict, and missing subsets. Through extensive evaluation, we uncover a Video Contribution Collapse (VCC) phenomenon, where MLLM marginalize video evidence due to high token redundancy and modality preferences. To address this, we propose Conflict-aware Head-level Attention Steering (CHASE), a lightweight mechanism that detects modality conflicts and performs inference-time attention steering, effectively mitigating decision bias without retraining the backbone. Experimental results demonstrate that CHASE consistently improves performance across various settings, significantly enhancing the reliability of MLLM in complex affective scenarios.

0 citations0 reviewsNo code linkedOpen access
preprint2025arXiv

MiMo-Audio: Audio Language Models are Few-Shot Learners

Existing audio language models typically rely on task-specific fine-tuning to accomplish particular audio tasks. In contrast, humans are able to generalize to new audio tasks with only a few examples or simple instructions. GPT-3 has shown that scaling next-token prediction pretraining enables strong generalization capabilities in text, and we believe this paradigm is equally applicable to the audio domain. By scaling MiMo-Audio's pretraining data to over one hundred million of hours, we observe the emergence of few-shot learning capabilities across a diverse set of audio tasks. We develop a systematic evaluation of these capabilities and find that MiMo-Audio-7B-Base achieves SOTA performance on both speech intelligence and audio understanding benchmarks among open-source models. Beyond standard metrics, MiMo-Audio-7B-Base generalizes to tasks absent from its training data, such as voice conversion, style transfer, and speech editing. MiMo-Audio-7B-Base also demonstrates powerful speech continuation capabilities, capable of generating highly realistic talk shows, recitations, livestreaming and debates. At the post-training stage, we curate a diverse instruction-tuning corpus and introduce thinking mechanisms into both audio understanding and generation. MiMo-Audio-7B-Instruct achieves open-source SOTA on audio understanding benchmarks (MMSU, MMAU, MMAR, MMAU-Pro), spoken dialogue benchmarks (Big Bench Audio, MultiChallenge Audio) and instruct-TTS evaluations, approaching or surpassing closed-source models. Model checkpoints and full evaluation suite are available at https://github.com/XiaomiMiMo/MiMo-Audio.

0 citations0 reviewsNo code linkedOpen access
preprint2024arXiv

Is Bigger and Deeper Always Better? Probing LLaMA Across Scales and Layers

This paper presents an in-depth analysis of Large Language Models (LLMs), focusing on LLaMA, a prominent open-source foundational model in natural language processing. Instead of assessing LLaMA through its generative output, we design multiple-choice tasks to probe its intrinsic understanding in high-order tasks such as reasoning and computation. We examine the model horizontally, comparing different sizes, and vertically, assessing different layers. We unveil several key and uncommon findings based on the designed probing tasks: (1) Horizontally, enlarging model sizes almost could not automatically impart additional knowledge or computational prowess. Instead, it can enhance reasoning abilities, especially in math problem solving, and helps reduce hallucinations, but only beyond certain size thresholds; (2) In vertical analysis, the lower layers of LLaMA lack substantial arithmetic and factual knowledge, showcasing logical thinking, multilingual and recognitive abilities, with top layers housing most computational power and real-world knowledge.

0 citations0 reviewsNo code linkedOpen access
preprint2023arXiv

A selection of Hα emitters at z = 2.1-2.5 using the Ks-band photometry of ZFOURGE

Large and less-biased samples of star-forming galaxies are essential to investigate galaxy evolution. H$\rmα$ emission line is one of the most reliable tracers of star-forming galaxies because its strength is directly related to recent star formation. However, it is observationally expensive to construct large samples of H$\rmα$ emitters by spectroscopic or narrow-band imaging survey at high-redshifts. In this work, we demonstrate a method to extract H$\rmα$ fluxes of galaxies at $z=2.1$-$2.5$ from $K_s$ broad-band photometry of ZFOURGE catalog. Combined with 25-39 other filters, we estimate the emission line fluxes by SED fitting with stellar population models that incorporate emission-line strengths. 2005 galaxies are selected as H$\rmα$ emitters by our method and their fluxes show good agreement with previous measurements in the literature. On the other hand, there are more H$\rmα$ luminous galaxies than previously reported. The discrepancy can be explained by extended H$\rmα$ profiles of massive galaxies and a luminosity dependence of dust attenuation, which are not taken into account in the previous work. We also find that there are a large number of low-mass galaxies with much higher specific star formation rate (sSFR) than expected from the extrapolated star formation main sequence. Such low-mass galaxies exhibit larger ratios between H$\rmα$ and UV fluxes compared to more massive high sSFR galaxies. This result implies that a ``starburst'' mode may differ among galaxies: low-mass galaxies appear to assemble their stellar mass via short-duration bursts while more massive galaxies tend to experience longer-duration ($>10\ \mathrm{Myr}$) bursts.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Automatic Prosody Annotation with Pre-Trained Text-Speech Model

Prosodic boundary plays an important role in text-to-speech synthesis (TTS) in terms of naturalness and readability. However, the acquisition of prosodic boundary labels relies on manual annotation, which is costly and time-consuming. In this paper, we propose to automatically extract prosodic boundary labels from text-audio data via a neural text-speech model with pre-trained audio encoders. This model is pre-trained on text and speech data separately and jointly fine-tuned on TTS data in a triplet format: {speech, text, prosody}. The experimental results on both automatic evaluation and human evaluation demonstrate that: 1) the proposed text-speech prosody annotation framework significantly outperforms text-only baselines; 2) the quality of automatic prosodic boundary annotations is comparable to human annotations; 3) TTS systems trained with model-annotated boundaries are slightly better than systems that use manual ones.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Bridging the Gap between Language Models and Cross-Lingual Sequence Labeling

Large-scale cross-lingual pre-trained language models (xPLMs) have shown effectiveness in cross-lingual sequence labeling tasks (xSL), such as cross-lingual machine reading comprehension (xMRC) by transferring knowledge from a high-resource language to low-resource languages. Despite the great success, we draw an empirical observation that there is a training objective gap between pre-training and fine-tuning stages: e.g., mask language modeling objective requires local understanding of the masked token and the span-extraction objective requires global understanding and reasoning of the input passage/paragraph and question, leading to the discrepancy between pre-training and xMRC. In this paper, we first design a pre-training task tailored for xSL named Cross-lingual Language Informative Span Masking (CLISM) to eliminate the objective gap in a self-supervised manner. Second, we present ContrAstive-Consistency Regularization (CACR), which utilizes contrastive learning to encourage the consistency between representations of input parallel sequences via unsupervised cross-lingual instance-wise training signals during pre-training. By these means, our methods not only bridge the gap between pretrain-finetune, but also enhance PLMs to better capture the alignment between different languages. Extensive experiments prove that our method achieves clearly superior results on multiple xSL benchmarks with limited pre-training data. Our methods also surpass the previous state-of-the-art methods by a large margin in few-shot data settings, where only a few hundred training examples are available.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Design optimization of band-pass filter based on parity-time symmetry coupled-resonant

Integrated optical filter based on microring resonators plays a critical role in many applications, ranging from wavelength division multiplexing and switching to channel routing. Bandwidth tunable filters are capable of meeting the on-demand flexible operations in complex situations, due to their advantages of scalability, multi-function, and energy-saving. It has been investigated recently that parity-time (PT) symmetry coupled-resonant systems can be applied to the bandwidth-tunable filters. However, due to the trade-off between the bandwidth-tunable contrast ratio and insertion loss of system, the bandwidth-tunable contrast ratio of this method is severely limited. Here, the bandwidth-tunable contrast ratio is defined as the maximum bandwidth divided by the minimum bandwidth. In this work, we show that high bandwidth-tunable contrast ratio and low insertion loss of system can be achieved simultaneously by increasing the coupling strength between the input port and the resonant. System characterizations under different coupling states reveal that the low insertion loss can be obtained when the system initially operates at the over-coupling condition. A high bandwidth-tunable contrast ratio PT-symmetry band-pass filter with moderate insertion loss is shown on the Silicon platform. Our scheme provides an effective method to reduce the insertion loss of on-chip tunable filters, which is also applicable to the high-order cascaded microring systems.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Detecting Recolored Image by Spatial Correlation

Image forensics, aiming to ensure the authenticity of the image, has made great progress in dealing with common image manipulation such as copy-move, splicing, and inpainting in the past decades. However, only a few researchers pay attention to an emerging editing technique called image recoloring, which can manipulate the color values of an image to give it a new style. To prevent it from being used maliciously, the previous approaches address the conventional recoloring from the perspective of inter-channel correlation and illumination consistency. In this paper, we try to explore a solution from the perspective of the spatial correlation, which exhibits the generic detection capability for both conventional and deep learning-based recoloring. Through theoretical and numerical analysis, we find that the recoloring operation will inevitably destroy the spatial correlation between pixels, implying a new prior of statistical discriminability. Based on such fact, we generate a set of spatial correlation features and learn the informative representation from the set via a convolutional neural network. To train our network, we use three recoloring methods to generate a large-scale and high-quality data set. Extensive experimental results in two recoloring scenes demonstrate that the spatial correlation features are highly discriminative. Our method achieves the state-of-the-art detection accuracy on multiple benchmark datasets and exhibits well generalization for unknown types of recoloring methods.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

End-to-end Spoken Conversational Question Answering: Task, Dataset and Model

In spoken question answering, the systems are designed to answer questions from contiguous text spans within the related speech transcripts. However, the most natural way that human seek or test their knowledge is via human conversations. Therefore, we propose a new Spoken Conversational Question Answering task (SCQA), aiming at enabling the systems to model complex dialogue flows given the speech documents. In this task, our main objective is to build the system to deal with conversational questions based on the audio recordings, and to explore the plausibility of providing more cues from different modalities with systems in information gathering. To this end, instead of directly adopting automatically generated speech transcripts with highly noisy data, we propose a novel unified data distillation approach, DDNet, which effectively ingests cross-modal information to achieve fine-grained representations of the speech and language modalities. Moreover, we propose a simple and novel mechanism, termed Dual Attention, by encouraging better alignments between audio and text to ease the process of knowledge transfer. To evaluate the capacity of SCQA systems in a dialogue-style interaction, we assemble a Spoken Conversational Question Answering (Spoken-CoQA) dataset with more than 40k question-answer pairs from 4k conversations. The performance of the existing state-of-the-art methods significantly degrade on our dataset, hence demonstrating the necessity of cross-modal information integration. Our experimental results demonstrate that our proposed method achieves superior performance in spoken conversational question answering tasks.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Exploring and Exploiting Multi-Granularity Representations for Machine Reading Comprehension

Recently, the attention-enhanced multi-layer encoder, such as Transformer, has been extensively studied in Machine Reading Comprehension (MRC). To predict the answer, it is common practice to employ a predictor to draw information only from the final encoder layer which generates the coarse-grained representations of the source sequences, i.e., passage and question. The analysis shows that the representation of source sequence becomes more coarse-grained from finegrained as the encoding layer increases. It is generally believed that with the growing number of layers in deep neural networks, the encoding process will gather relevant information for each location increasingly, resulting in more coarse-grained representations, which adds the likelihood of similarity to other locations (referring to homogeneity). Such phenomenon will mislead the model to make wrong judgement and degrade the performance. In this paper, we argue that it would be better if the predictor could exploit representations of different granularity from the encoder, providing different views of the source sequences, such that the expressive power of the model could be fully utilized. To this end, we propose a novel approach called Adaptive Bidirectional Attention-Capsule Network (ABA-Net), which adaptively exploits the source representations of different levels to the predictor. Furthermore, due to the better representations are at the core for boosting MRC performance, the capsule network and self-attention module are carefully designed as the building blocks of our encoders, which provides the capability to explore the local and global representations, respectively. Experimental results on three benchmark datasets, i.e., SQuAD 1.0, SQuAD 2.0 and COQA, demonstrate the effectiveness of our approach. In particular, we set the new state-of-the-art performance on the SQuAD 1.0 dataset

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

On Performance Loss of DOA Measurement Using Massive MIMO Receiver with Mixed-ADCs

High hardware cost and high power consumption of massive multiple-input and multiple output (MIMO) are two challenges for the future wireless communications including beyond fifth generation (B5G) and sixth generation (6G). Adopting the low-resolution analog-to-digital converter (ADC) is viewed as a promising solution. Additionally, the direction of arrival (DOA) estimation is an indispensable technology for beam alignment and tracking in massive MIMO systems. Thus, in this paper, the performance of DOA estimation with mixed-ADC structure is firstly investigated. The Cramer-Rao lower bound (CRLB) for this architecture is derived based on the additive quantization noise model. Eventually, a performance loss factor and the associated energy efficiency factor is defined for analysis in detail. Simulation results show that the mixed-ADC architecture can strike a good balance among performance loss, circuit cost and energy efficiency. More importantly, just a few bits (up to 4 bits) of low-resolution ADCs can achieve a satisfactory performance for DOA measurement.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Adaptive Bi-directional Attention: Exploring Multi-Granularity Representations for Machine Reading Comprehension

Recently, the attention-enhanced multi-layer encoder, such as Transformer, has been extensively studied in Machine Reading Comprehension (MRC). To predict the answer, it is common practice to employ a predictor to draw information only from the final encoder layer which generates the \textit{coarse-grained} representations of the source sequences, i.e., passage and question. Previous studies have shown that the representation of source sequence becomes more \textit{coarse-grained} from \textit{fine-grained} as the encoding layer increases. It is generally believed that with the growing number of layers in deep neural networks, the encoding process will gather relevant information for each location increasingly, resulting in more \textit{coarse-grained} representations, which adds the likelihood of similarity to other locations (referring to homogeneity). Such a phenomenon will mislead the model to make wrong judgments so as to degrade the performance. To this end, we propose a novel approach called Adaptive Bidirectional Attention, which adaptively exploits the source representations of different levels to the predictor. Experimental results on the benchmark dataset, SQuAD 2.0 demonstrate the effectiveness of our approach, and the results are better than the previous state-of-the-art model by 2.5$\%$ EM and 2.3$\%$ F1 scores.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Impact of Low-Resolution ADC on DOA Estimation Performance for Massive MIMO Receive Array

In this paper, we present a new scenario of direction of arrival (DOA) estimation using massive multiple-input multiple-output (MIMO) receive array with low-resolution analog-to-digital convertors (ADCs), which can strike a good balance between performance and circuit cost. Based on the linear additive quantization noise model (AQNM), the effect of low-resolution ADCs on the methods, such as Root-MUSIC method, is analyzed. Also, the closed-form expression of Cramer-Rao lower bound (CRLB) is derived to evaluate the performance loss caused by the low-resolution ADCs. The simulation results show that the Root-MUSIC methods can achieve the corresponding CRLB. Furthermore, 2-3 bits are acceptable for most applications if the 1dB performance loss.

0 citations0 reviewsNo code linkedOpen access