arXiv subjects

Jianqing Sun

Publications and source records attributed to Jianqing Sun.

7 recordsLinked to original sources

arXiv2026

SoftRerank: Hierarchical Soft Fusion with Candidate-Label Reranking for Long-Tailed Micro-Action Recognition

Micro-actions are subtle, low-intensity non-verbal behaviors that provide cues to fine-grained human states, including emotions and intentions. Recognizing them remains difficult because they are brief, contain weak visual changes, and often exhibit similar motion patterns across categories. This paper addresses these challenges with a fine-grained micro-action recognition method that combines full fine-tuning of InternVideo2.5, hierarchical soft fusion, and a lightweight candidate-label reranker. For the long-tailed label distribution in MA-52, we use class-balanced sampling and inverse-frequency reweighting to reduce the effect of frequent classes during training. We fine-tune InternVideo2.5 end to end and attach coarse and group-conditional fine-grained classification heads to the shared video representation, improving the consistency between coarse and fine predictions. For ambiguous samples, the candidate-label reranker uses hard samples and video-label matching to focus on easily confused fine-grained actions. Experiments validate the proposed method, which achieves a 79.99% F1-mean on MA-52 and ranks first in the 3rd Micro-Action Analysis Grand Challenge at ACM Multimedia 2026.

cs.CV↗

arXiv2026

Semi-Supervised Facial Expression Recognition based on Dynamic Threshold and Negative Learning

Facial expression recognition is a key task in human-computer interaction and affective computing. However, acquiring a large amount of labeled facial expression data is often costly. Therefore, it is particularly important to design a semi-supervised facial expression recognition algorithm that makes full use of both labeled and unlabeled data. In this paper, we propose a semi-supervised facial expression recognition algorithm based on Dynamic Threshold Adjustment (DTA) and Selective Negative Learning (SNL). Initially, we designed strategies for local attention enhancement and random dropout of feature maps during feature extraction, which strengthen the representation of local features while ensuring the model does not overfit to any specific local area. Furthermore, this study introduces a dynamic thresholding method to adapt to the requirements of the semi-supervised learning framework for facial expression recognition tasks, and through a selective negative learning strategy, it fully utilizes unlabeled samples with low confidence by mining useful expression information from complementary labels, achieving impressive results. We have achieved state-of-the-art performance on the RAF-DB and AffectNet datasets. Our method surpasses fully supervised methods even without using the entire dataset, which proves the effectiveness of our approach.

cs.CV↗

arXiv2023

Rhythm-controllable Attention with High Robustness for Long Sentence Speech Synthesis

Regressive Text-to-Speech (TTS) system utilizes attention mechanism to generate alignment between text and acoustic feature sequence. Alignment determines synthesis robustness (e.g, the occurence of skipping, repeating, and collapse) and rhythm via duration control. However, current attention algorithms used in speech synthesis cannot control rhythm using external duration information to generate natural speech while ensuring robustness. In this study, we propose Rhythm-controllable Attention (RC-Attention) based on Tracotron2, which improves robustness and naturalness simultaneously. Proposed attention adopts a trainable scalar learned from four kinds of information to achieve rhythm control, which makes rhythm control more robust and natural, even when synthesized sentences are extremely longer than training corpus. We use word errors counting and AB preference test to measure robustness of proposed method and naturalness of synthesized speech, respectively. Results shows that RC-Attention has the lowest word error rate of nearly 0.6%, compared with 11.8% for baseline system. Moreover, nearly 60% subjects prefer to the speech synthesized with RC-Attention to that with Forward Attention, because the former has more natural rhythm.

cs.AI↗

arXiv2023

M2-CTTS: End-to-End Multi-scale Multi-modal Conversational Text-to-Speech Synthesis

Conversational text-to-speech (TTS) aims to synthesize speech with proper prosody of reply based on the historical conversation. However, it is still a challenge to comprehensively model the conversation, and a majority of conversational TTS systems only focus on extracting global information and omit local prosody features, which contain important fine-grained information like keywords and emphasis. Moreover, it is insufficient to only consider the textual features, and acoustic features also contain various prosody information. Hence, we propose M2-CTTS, an end-to-end multi-scale multi-modal conversational text-to-speech system, aiming to comprehensively utilize historical conversation and enhance prosodic expression. More specifically, we design a textual context module and an acoustic context module with both coarse-grained and fine-grained modeling. Experimental results demonstrate that our model mixed with fine-grained context information and additionally considering acoustic features achieves better prosody performance and naturalness in CMOS tests.

cs.SD↗

arXiv2022

ECAPA-TDNN for Multi-speaker Text-to-speech Synthesis

In recent years, neural network based methods for multi-speaker text-to-speech synthesis (TTS) have made significant progress. However, the current speaker encoder models used in these methods still cannot capture enough speaker information. In this paper, we focus on accurate speaker encoder modeling and propose an end-to-end method that can generate high-quality speech and better similarity for both seen and unseen speakers. The proposed architecture consists of three separately trained components: a speaker encoder based on the state-of-the-art ECAPA-TDNN model which is derived from speaker verification task, a FastSpeech2 based synthesizer, and a HiFi-GAN vocoder. The comparison among different speaker encoder models shows our proposed method can achieve better naturalness and similarity. To efficiently evaluate our synthesized speech, we are the first to adopt deep learning based automatic MOS evaluation methods to assess our results, and these methods show great potential in automatic speech quality assessment.

cs.SD↗

arXiv2021

Resolving Two Distinct Thermal X-ray Components in A compound Solar Flare

X-ray emission provides the most direct diagnostics of the energy-release process in solar flares. Occasionally, a superhot X-ray source is found to be above hot flare loops of ~10 MK temperature. While the origin of the superhot plasma is still elusive, it has conjured up an intriguing image of in-situ plasma heating near the reconnection site high above the flare loops, in contrast to the conventional picture of chromospheric evaporation. Here we investigate an extremely long-duration solar flare, in which EUV images show two distinct flare loop systems that appear successively along a Gamma-shaped polarity inversion line (PIL). When both flare loop systems are present, the HXR spectrum is found to be well fitted by combining a hot component (Te ~12 MK) and a superhot component (Te ~30 MK). Associated with a fast CME, the superhot X-ray source is located at top of the flare arcade that appears earlier, straddling and extending along the long "arm" of the Gamma-shaped PIL. Associated with a slow CME, the hot X-ray source is located at the top of the flare arcade that appears later and sits astride the short "arm" of the Gamma-shaped PIL. Aided by observations from a different viewing angle, we are able to verify that the superhot X-ray source is above the hot one in projection, but the two sources belong to different flare loop systems. Thus, this case study provides a stereoscopic observation explaining the co-existence of superhot and hot X-ray emitting plasmas in solar flares.

astro-ph.SR↗

arXiv2017

Numerical calculation of N-periodic wave solutions to coupled KdV-type equations

In this paper, we study periodic wave solutions of coupled KdV-type equations. We present a numerical process to calculate the $N$-periodic waves based on the direct method of calculating periodic wave solutions proposed by Akira Nakamura. Particularly, in the case of $N=1,2,3$, we give some detailed examples to show the N-periodic wave solutions to a coupled Ramani equation.

math-ph↗