arXiv subjects

Feiran Yang

Publications and source records attributed to Feiran Yang.

8 recordsLinked to original sources

arXiv2025

Improving Efficiency in Federated Learning with Optimized Homomorphic Encryption

Federated learning is a method used in machine learning to allow multiple devices to work together on a model without sharing their private data. Each participant keeps their private data on their system and trains a local model and only sends updates to a central server, which combines these updates to improve the overall model. A key enabler of privacy in FL is homomorphic encryption (HE). HE allows computations to be performed directly on encrypted data. While HE offers strong privacy guarantees, it is computationally intensive, leading to significant latency and scalability issues, particularly for large-scale models like BERT. In my research, I aimed to address this inefficiency problem. My research introduces a novel algorithm to address these inefficiencies while maintaining robust privacy guarantees. I integrated several mathematical techniques such as selective parameter encryption, sensitivity maps, and differential privacy noise within my algorithms, which has already improved its efficiency. I have also conducted rigorous mathematical proofs to validate the correctness and robustness of the approach. I implemented this algorithm by coding it in C++, simulating the environment of federated learning on large-scale models, and verified that the efficiency of my algorithm is $3$ times the efficiency of the state-of-the-art method. This research has significant implications for machine learning because its ability to improve efficiency while balancing privacy makes it a practical solution! It would enable federated learning to be used very efficiently and deployed in various resource-constrained environments, as this research provides a novel solution to one of the key challenges in federated learning: the inefficiency of homomorphic encryption, as my new algorithm is able to enhance the scalability and resource efficiency of FL while maintaining robust privacy guarantees.

cs.CR↗

arXiv2024

PSELDNets: Pre-trained Neural Networks on a Large-scale Synthetic Dataset for Sound Event Localization and Detection

Sound event localization and detection (SELD) has seen substantial advancements through learning-based methods. These systems, typically trained from scratch on specific datasets, have shown considerable generalization capabilities. Recently, deep neural networks trained on large-scale datasets have achieved remarkable success in the sound event classification (SEC) field, prompting an open question of whether these advances can be extended to the development of SELD foundation models. In this paper, leveraging the power of pre-trained SEC models, we propose pre-trained SELD networks (PSELDNets) on a large-scale synthetic dataset. The synthetic dataset, generated by convolving sound events with simulated spatial room impulse responses (SRIRs), contains 1,167 hours of audio clips with an ontology of 170 sound classes. These PSELDNets are applied to various SELD scenarios. When we adapt PSELDNets to specific scenarios, particularly in cases of low-resource data, we introduce a data-efficient fine-tuning method, AdapterBit. PSELDNets are evaluated on synthetic-test-set using collected SRIRs from the TAU Spatial Room Impulse Response Database (TAU-SRIR DB) and achieve satisfactory performance. We also carried out experiments to validate the transferability of PSELDNets to three publicly available datasets and our own real-world recordings. The results demonstrate that PSELDNets surpass state-of-the-art systems across all publicly available datasets. Given the need for direction-of-arrival estimation, SELD generally relies on sufficient multi-channel audio clips. However, incorporating the AdapterBit, PSELDNets show more efficient adaptability to various scenarios using minimal multi-channel or even just monophonic audio clips, outperforming traditional fine-tuning approaches.

eess.AS↗

arXiv2023

Selective-Memory Meta-Learning with Environment Representations for Sound Event Localization and Detection

Environment shifts and conflicts present significant challenges for learning-based sound event localization and detection (SELD) methods. SELD systems, when trained in particular acoustic settings, often show restricted generalization capabilities for diverse acoustic environments. Furthermore, obtaining annotated samples for spatial sound events is notably costly. Deploying a SELD system in a new environment requires extensive time for re-training and fine-tuning. To overcome these challenges, we propose environment-adaptive Meta-SELD, designed for efficient adaptation to new environments using minimal data. Our method specifically utilizes computationally synthesized spatial data and employs Model-Agnostic Meta-Learning (MAML) on a pre-trained, environment-independent model. The method then utilizes fast adaptation to unseen real-world environments using limited samples from the respective environments. Inspired by the Learning-to-Forget approach, we introduce the concept of selective memory as a strategy for resolving conflicts across environments. This approach involves selectively memorizing target-environment-relevant information and adapting to the new environments through the selective attenuation of model parameters. In addition, we introduce environment representations to characterize different acoustic settings, enhancing the adaptability of our attenuation approach to various environments. We evaluate our proposed method on the development set of the Sony-TAu Realistic Spatial Soundscapes 2023 (STARSS23) dataset and computationally synthesized scenes. Experimental results demonstrate the superior performance of the proposed method compared to conventional supervised learning methods, particularly in localization.

eess.AS↗

arXiv2023

META-SELD: Meta-Learning for Fast Adaptation to the new environment in Sound Event Localization and Detection

For learning-based sound event localization and detection (SELD) methods, different acoustic environments in the training and test sets may result in large performance differences in the validation and evaluation stages. Different environments, such as different sizes of rooms, different reverberation times, and different background noise, may be reasons for a learning-based system to fail. On the other hand, acquiring annotated spatial sound event samples, which include onset and offset time stamps, class types of sound events, and direction-of-arrival (DOA) of sound sources is very expensive. In addition, deploying a SELD system in a new environment often poses challenges due to time-consuming training and fine-tuning processes. To address these issues, we propose Meta-SELD, which applies meta-learning methods to achieve fast adaptation to new environments. More specifically, based on Model Agnostic Meta-Learning (MAML), the proposed Meta-SELD aims to find good meta-initialized parameters to adapt to new environments with only a small number of samples and parameter updating iterations. We can then quickly adapt the meta-trained SELD model to unseen environments. Our experiments compare fine-tuning methods from pre-trained SELD models with our Meta-SELD on the Sony-TAU Realistic Spatial Soundscapes 2023 (STARSSS23) dataset. The evaluation results demonstrate the effectiveness of Meta-SELD when adapting to new environments.

eess.AS↗

arXiv2022

Sound Event Localization and Detection for Real Spatial Sound Scenes: Event-Independent Network and Data Augmentation Chains

Sound event localization and detection (SELD) is a joint task of sound event detection and direction-of-arrival estimation. In DCASE 2022 Task 3, types of data transform from computationally generated spatial recordings to recordings of real-sound scenes. Our system submitted to the DCASE 2022 Task 3 is based on our previous proposed Event-Independent Network V2 (EINV2) with a novel data augmentation method. Our method employs EINV2 with a track-wise output format, permutation-invariant training, and a soft parameter-sharing strategy, to detect different sound events of the same class but in different locations. The Conformer structure is used for extending EINV2 to learn local and global features. A data augmentation method, which contains several data augmentation chains composed of stochastic combinations of several different data augmentation operations, is utilized to generalize the model. To mitigate the lack of real-scene recordings in the development dataset and the presence of sound events being unbalanced, we exploit FSD50K, AudioSet, and TAU Spatial Room Impulse Response Database (TAU-SRIR DB) to generate simulated datasets for training. We present results on the validation set of Sony-TAu Realistic Spatial Soundscapes 2022 (STARSS22) in detail. Experimental results indicate that the ability to generalize to different environments and unbalanced performance among different classes are two main challenges. We evaluate our proposed method in Task 3 of the DCASE 2022 challenge and obtain the second rank in the teams ranking. Source code is released.

eess.AS↗

arXiv2022

A Track-Wise Ensemble Event Independent Network for Polyphonic Sound Event Localization and Detection

Polyphonic sound event localization and detection (SELD) aims at detecting types of sound events with corresponding temporal activities and spatial locations. In this paper, a track-wise ensemble event independent network with a novel data augmentation method is proposed. The proposed model is based on our previous proposed Event-Independent Network V2 and is extended by conformer blocks and dense blocks. The track-wise ensemble model with track-wise output format is proposed to solve an ensemble model problem for track-wise output format that track permutation may occur among different models. The data augmentation approach contains several data augmentation chains, which are composed of random combinations of several data augmentation operations. The method also utilizes log-mel spectrograms, intensity vectors, and Spatial Cues-Augmented Log-Spectrogram (SALSA) for different models. We evaluate our proposed method in the Task of the L3DAS22 challenge and obtain the top ranking solution with a location-dependent F-score to be 0.699. Source code is released.

cs.SD↗

arXiv2020

Colourings of $(m, n)$-coloured mixed graphs

A mixed graph is, informally, an object obtained from a simple undirected graph by choosing an orientation for a subset of its edges. A mixed graph is $(m, n)$-coloured if each edge is assigned one of $m \geq 0$ colours, and each arc is assigned one of $n \geq 0$ colours. Oriented graphs are $(0, 1)$-coloured mixed graphs, and 2-edge-coloured graphs are $(2, 0)$-coloured mixed graphs. We show that results of Sopena for vertex colourings of oriented graphs, and of Kostochka, Sopena and Zhu for vertex colourings oriented graphs and 2-edge-coloured graphs, are special cases of results about vertex colourings of $(m, n)$-coloured mixed graphs. Both of these can be regarded as a version of Brooks' Theorem.

math.CO↗

arXiv2017

Complexity of locally-injective homomorphisms to tournaments

For oriented graphs $G$ and $H$, a homomorphism $f: G \rightarrow H$ is locally-injective if, for every $v \in V(G)$, it is injective when restricted to some combination of the in-neighbourhood and out-neighbourhood of $v$. Two of the possible definitions of local-injectivity are examined. In each case it is shown that the associated homomorphism problem is NP-complete when $H$ is a reflexive tournament on three or more vertices with a loop at every vertex, and solvable in polynomial time when $H$ is a reflexive tournament on two or fewer vertices.

cs.DM↗