Author name cluster

Yan Fan

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

3 papers

1 author row

AAAI Conference 2024 Conference Paper

Dynamic Sub-graph Distillation for Robust Semi-supervised Continual Learning

Yan Fan
Yu Wang
Pengfei Zhu
Qinghua Hu

Continual learning (CL) has shown promising results and comparable performance to learning at once in a fully supervised manner. However, CL strategies typically require a large number of labeled samples, making their real-life deployment challenging. In this work, we focus on semi-supervised continual learning (SSCL), where the model progressively learns from partially labeled data with unknown categories. We provide a comprehensive analysis of SSCL and demonstrate that unreliable distributions of unlabeled data lead to unstable training and refinement of the progressing stages. This problem severely impacts the performance of SSCL. To address the limitations, we propose a novel approach called Dynamic Sub-Graph Distillation (DSGD) for semi-supervised continual learning, which leverages both semantic and structural information to achieve more stable knowledge distillation on unlabeled data and exhibit robustness against distribution bias. Firstly, we formalize a general model of structural distillation and design a dynamic graph construction for the continual learning progress. Next, we define a structure distillation vector and design a dynamic sub-graph distillation algorithm, which enables end-to-end training and adaptability to scale up tasks. The entire proposed method is adaptable to various CL methods and supervision settings. Finally, experiments conducted on three datasets CIFAR10, CIFAR100, and ImageNet-100, with varying supervision ratios, demonstrate the effectiveness of our proposed approach in mitigating the catastrophic forgetting problem in semi-supervised continual learning scenarios. Our code is available: https://github.com/fanyan0411/DSGD.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Persistence Homology Distillation for Semi-supervised Continual Learning

Yan Fan
Yu Wang
Pengfei Zhu
Dongyue Chen
Qinghua Hu

Semi-supervised continual learning (SSCL) has attracted significant attention for addressing catastrophic forgetting in semi-supervised data. Knowledge distillation, which leverages data representation and pair-wise similarity, has shown significant potential in preserving information in SSCL. However, traditional distillation strategies often fail in unlabeled data with inaccurate or noisy information, limiting their efficiency in feature spaces undergoing substantial changes during continual learning. To address these limitations, we propose Persistence Homology Distillation (PsHD) to preserve intrinsic structural information that is insensitive to noise in semi-supervised continual learning. First, we capture the structural features using persistence homology by homological evolution across different scales in vision data, where the multi-scale characteristic established its stability under noise interference. Next, we propose a persistence homology distillation loss in SSCL and design an acceleration algorithm to reduce the computational cost of persistence homology in our module. Furthermore, we demonstrate the superior stability of PsHD compared to sample representation and pair-wise similarity distillation methods theoretically and experimentally. Finally, experimental results on three widely used datasets validate that the new PsHD outperforms state-of-the-art with 3. 9% improvements on average, and also achieves 1. 5% improvements while reducing 60% memory buffer size, highlighting the potential of utilizing unlabeled data in SSCL. Our code is available: https: //github. com/fanyan0411/PsHD.

PDF Details DOI

YNIMG Journal 2019 Journal Article

The anterior insula channels prefrontal expectancy signals during affective processing

Vanessa Teckentrup
Johan N. van der Meer
Viola Borchardt
Yan Fan
Monja P. Neuser
Claus Tempelmann
Luisa Herrmann
Martin Walter

Expectancy shapes our perception of impending events. Although such an interplay between cognitive and affective processes is often impaired in mental disorders, it is not well understood how top-down expectancy signals modulate future affect. We therefore track the information flow in the brain during cognitive and affective processing segregated in time using task-specific cross-correlations. Participants in two independent fMRI studies (N1 = 37 & N2 = 55) were instructed to imagine a situation with affective content as indicated by a cue, which was then followed by an emotional picture congruent with expectancy. To correct for intrinsic covariance of brain function, we calculate resting-state cross-correlations analogous to the task. First, using factorial modeling of delta cross-correlations (task-rest) of the first study, we find that the magnitude of expectancy signals in the anterior insula cortex (AIC) modulates the BOLD response to emotional pictures in the anterior cingulate and dorsomedial prefrontal cortex in opposite directions. Second, using hierarchical linear modeling of lagged connectivity, we demonstrate that expectancy signals in the AIC indeed foreshadow this opposing pattern in the prefrontal cortex. Third, we replicate the results in the second study using a higher temporal resolution, showing that our task-specific cross-correlation approach robustly uncovers the dynamics of information flow. We conclude that the AIC arbitrates the recruitment of distinct prefrontal networks during cued picture processing according to triggered expectations. Taken together, our study provides new insights into neuronal pathways channeling cognition and affect within well-defined brain networks. Better understanding of such dynamics could lead to new applications tracking aberrant information processing in mental disorders.

Details DOI