Saeed Vahidian

Saeed Vahidian, Mahdi Morafah, Weijia Wang et al.

Clustered federated learning (FL) has been shown to produce promising results by grouping clients into clusters. This is especially effective in scenarios where separate groups of clients have significant differences in the distributions of their local data. Existing clustered FL algorithms are essentially trying to group together clients with similar distributions so that clients in the same cluster can leverage each other's data to better perform federated learning. However, prior clustered FL algorithms attempt to learn these distribution similarities indirectly during training, which can be quite time consuming as many rounds of federated learning may be required until the formation of clusters is stabilized. In this paper, we propose a new approach to federated learning that directly aims to efficiently identify distribution similarities among clients by analyzing the principal angles between the client data subspaces. Each client applies a truncated singular value decomposition (SVD) step on its local data in a single-shot manner to derive a small set of principal vectors, which provides a signature that succinctly captures the main characteristics of the underlying distribution. This small set of principal vectors is provided to the server so that the server can directly identify distribution similarities among the clients to form clusters. This is achieved by comparing the similarities of the principal angles between the client data subspaces spanned by those principal vectors. The approach provides a simple, yet effective clustered FL framework that addresses a broad range of data heterogeneity issues beyond simpler forms of Non-IIDness like label skews. Our clustered FL approach also enables convergence guarantees for non-convex objectives. Our code is available at https://github.com/MMorafah/PACFL.

Mahdi Morafah, Saeed Vahidian, Chen Chen et al.

Though successful, federated learning presents new challenges for machine learning, especially when the issue of data heterogeneity, also known as Non-IID data, arises. To cope with the statistical heterogeneity, previous works incorporated a proximal term in local optimization or modified the model aggregation scheme at the server side or advocated clustered federated learning approaches where the central server groups agent population into clusters with jointly trainable data distributions to take the advantage of a certain level of personalization. While effective, they lack a deep elaboration on what kind of data heterogeneity and how the data heterogeneity impacts the accuracy performance of the participating clients. In contrast to many of the prior federated learning approaches, we demonstrate not only the issue of data heterogeneity in current setups is not necessarily a problem but also in fact it can be beneficial for the FL participants. Our observations are intuitive: (1) Dissimilar labels of clients (label skew) are not necessarily considered data heterogeneity, and (2) the principal angle between the agents' data subspaces spanned by their corresponding principal vectors of data is a better estimate of the data heterogeneity. Our code is available at https://github.com/MMorafah/FL-SC-NIID.

Jianyang Gu, Saeed Vahidian, Vyacheslav Kungurtsev et al.

Dataset distillation reduces the storage and computational consumption of training a network by generating a small surrogate dataset that encapsulates rich information of the original large-scale one. However, previous distillation methods heavily rely on the sample-wise iterative optimization scheme. As the images-per-class (IPC) setting or image resolution grows larger, the necessary computation will demand overwhelming time and resources. In this work, we intend to incorporate generative diffusion techniques for computing the surrogate dataset. Observing that key factors for constructing an effective surrogate dataset are representativeness and diversity, we design additional minimax criteria in the generative training to enhance these facets for the generated images of diffusion models. We present a theoretical model of the process as hierarchical diffusion control demonstrating the flexibility of the diffusion process to target these criteria without jeopardizing the faithfulness of the sample to the desired distribution. The proposed method achieves state-of-the-art validation performance while demanding much less computational resources. Under the 100-IPC setting on ImageWoof, our method requires less than one-twentieth the distillation time of previous methods, yet yields even better performance. Source code and generated data are available in https://github.com/vimar-gu/MinimaxDiffusion.

Mahdi Morafah, Saeed Vahidian, Weijia Wang et al.

Classical federated learning approaches yield significant performance degradation in the presence of Non-IID data distributions of participants. When the distribution of each local dataset is highly different from the global one, the local objective of each client will be inconsistent with the global optima which incur a drift in the local updates. This phenomenon highly impacts the performance of clients. This is while the primary incentive for clients to participate in federated learning is to obtain better personalized models. To address the above-mentioned issue, we present a new algorithm, FLIS, which groups the clients population in clusters with jointly trainable data distributions by leveraging the inference similarity of clients' models. This framework captures settings where different groups of users have their own objectives (learning tasks) but by aggregating their data with others in the same cluster (same learning task) to perform more efficient and personalized federated learning. We present experimental results to demonstrate the benefits of FLIS over the state-of-the-art benchmarks on CIFAR-100/10, SVHN, and FMNIST datasets. Our code is available at https://github.com/MMorafah/FLIS.

Umar Khalid, Hasan Iqbal, Saeed Vahidian et al.

Human-robot interaction (HRI) is a rapidly growing field that encompasses social and industrial applications. Machine learning plays a vital role in industrial HRI by enhancing the adaptability and autonomy of robots in complex environments. However, data privacy is a crucial concern in the interaction between humans and robots, as companies need to protect sensitive data while machine learning algorithms require access to large datasets. Federated Learning (FL) offers a solution by enabling the distributed training of models without sharing raw data. Despite extensive research on Federated learning (FL) for tasks such as natural language processing (NLP) and image classification, the question of how to use FL for HRI remains an open research problem. The traditional FL approach involves transmitting large neural network parameter matrices between the server and clients, which can lead to high communication costs and often becomes a bottleneck in FL. This paper proposes a communication-efficient FL framework for human-robot interaction (CEFHRI) to address the challenges of data heterogeneity and communication costs. The framework leverages pre-trained models and introduces a trainable spatiotemporal adapter for video understanding tasks in HRI. Experimental results on three human-robot interaction benchmark datasets: HRI30, InHARD, and COIN demonstrate the superiority of CEFHRI over full fine-tuning in terms of communication costs. The proposed methodology provides a secure and efficient approach to HRI federated learning, particularly in industrial environments with data privacy concerns and limited communication bandwidth. Our code is available at https://github.com/umarkhalidAI/CEFHRI-Efficient-Federated-Learning.

Saeed Vahidian, Sreevatsank Kadaveru, Woonjoon Baek et al.

An oft-cited open problem of federated learning is the existence of data heterogeneity at the clients. One pathway to understanding the drastic accuracy drop in federated learning is by scrutinizing the behavior of the clients' deep models on data with different levels of "difficulty", which has been left unaddressed. In this paper, we investigate a different and rarely studied dimension of FL: ordered learning. Specifically, we aim to investigate how ordered learning principles can contribute to alleviating the heterogeneity effects in FL. We present theoretical analysis and conduct extensive empirical studies on the efficacy of orderings spanning three kinds of learning: curriculum, anti-curriculum, and random curriculum. We find that curriculum learning largely alleviates non-IIDness. Interestingly, the more disparate the data distributions across clients the more they benefit from ordered learning. We provide analysis explaining this phenomenon, specifically indicating how curriculum training appears to make the objective landscape progressively less convex, suggesting fast converging iterations at the beginning of the training procedure. We derive quantitative results of convergence for both convex and nonconvex objectives by modeling the curriculum training on federated devices as local SGD with locally biased stochastic gradients. Also, inspired by ordered learning, we propose a novel client selection technique that benefits from the real-world disparity in the clients. Our proposed approach to client selection has a synergic effect when applied together with ordered learning in FL.

Dongyao Zhu, Zhen Wang, Xi Xiao et al.

Latent visual reasoning involves visual evidence more directly in multimodal reasoning by inserting continuous latent tokens before textual generation. However, the necessity of these latent tokens at inference remains ambiguous. We show that replacing latent tokens with random noise or removing them completely causes little performance degradation across spatial reasoning benchmarks. Reinforcement learning further diminishes the latent generation behavior after post-training. These observations raise a central question: Is latent visual reasoning still meaningful? We argue that its value should be measured by how effectively latent tokens guide learning, rather than whether they persist as an inference-time format. Our analysis shows that latent reasoning is unevenly favorable across question types, yet hard task-level routing for applying latent generation is brittle. Motivated by these findings, we propose an attention-based reward that encourages generated latent tokens to interact with later text tokens during RL. This reward promotes latent utilization when the latent mode is activated while preserving the flexibility to use pure-text reasoning. Experiments show that our method improves performance across perception and visual reasoning benchmarks, even when latent tokens are rarely generated after post-training. Our results highlight that, without explicit expression at inference, latent visual reasoning can shape better visual grounding and more accurate textual reasoning in silence. Our code and trained models are publicly available at \href{https://github.com/ddydyd32/silent-lvr/tree/master}{GitHub} and \href{https://huggingface.co/collections/cornuHGF/silent-lvr}{Hugging Face}.

Saeed Vahidian, Mingyu Wang, Jianyang Gu et al.

Dataset distillation (DD) has emerged as a widely adopted technique for crafting a synthetic dataset that captures the essential information of a training dataset, facilitating the training of accurate neural models. Its applications span various domains, including transfer learning, federated learning, and neural architecture search. The most popular methods for constructing the synthetic data rely on matching the convergence properties of training the model with the synthetic dataset and the training dataset. However, using the empirical loss as the criterion must be thought of as auxiliary in the same sense that the training set is an approximate substitute for the population distribution, and the latter is the data of interest. Yet despite its popularity, an aspect that remains unexplored is the relationship of DD to its generalization, particularly across uncommon subgroups. That is, how can we ensure that a model trained on the synthetic dataset performs well when faced with samples from regions with low population density? Here, the representativeness and coverage of the dataset become salient over the guaranteed training error at inference. Drawing inspiration from distributionally robust optimization, we introduce an algorithm that combines clustering with the minimization of a risk measure on the loss to conduct DD. We provide a theoretical rationale for our approach and demonstrate its effective generalization and robustness across subgroups through numerical experiments. The source code is available at https://github.com/Mming11/RobustDatasetDistillation.

Jicang Cai, Saeed Vahidian, Weijia Wang et al.

Meta-learning often referred to as learning-to-learn is a promising notion raised to mimic human learning by exploiting the knowledge of prior tasks but being able to adapt quickly to novel tasks. A plethora of models has emerged in this context and improved the learning efficiency, robustness, etc. The question that arises here is can we emulate other aspects of human learning and incorporate them into the existing meta learning algorithms? Inspired by the widely recognized finding in neuroscience that distinct parts of the brain are highly specialized for different types of tasks, we aim to improve the model performance of the current meta learning algorithms by selectively using only parts of the model conditioned on the input tasks. In this work, we describe an approach that investigates task-dependent dynamic neuron selection in deep convolutional neural networks (CNNs) by leveraging the scaling factor in the batch normalization (BN) layer associated with each convolutional layer. The problem is intriguing because the idea of helping different parts of the model to learn from different types of tasks may help us train better filters in CNNs, and improve the model generalization performance. We find that the proposed approach, neural routing in meta learning (NRML), outperforms one of the well-known existing meta learning baselines on few-shot classification tasks on the most widely used benchmark datasets.

Yao Lu, Jianyang Gu, Xuguang Chen et al.

Dataset Distillation (DD) is a promising technique to synthesize a smaller dataset that preserves essential information from the original dataset. This synthetic dataset can serve as a substitute for the original large-scale one, and help alleviate the training workload. However, current DD methods typically operate under the assumption that the dataset is unbiased, overlooking potential bias issues within the dataset itself. To fill in this blank, we systematically investigate the influence of dataset bias on DD. To the best of our knowledge, this is the first exploration in the DD domain. Given that there are no suitable biased datasets for DD, we first construct two biased datasets, CMNIST-DD and CCIFAR10-DD, to establish a foundation for subsequent analysis. Then we utilize existing DD methods to generate synthetic datasets on CMNIST-DD and CCIFAR10-DD, and evaluate their performance following the standard process. Experiments demonstrate that biases present in the original dataset significantly impact the performance of the synthetic dataset in most cases, which highlights the necessity of identifying and mitigating biases in the original datasets during DD. Finally, we reformulate DD within the context of a biased dataset. Our code along with biased datasets are available at https://github.com/yaolu-zjut/Biased-DD.

Vyacheslav Kungurtsev, Yuanfang Peng, Jianyang Gu et al.

Dataset distillation (DD) is an increasingly important technique that focuses on constructing a synthetic dataset capable of capturing the core information in training data to achieve comparable performance in models trained on the latter. While DD has a wide range of applications, the theory supporting it is less well evolved. New methods of DD are compared on a common set of benchmarks, rather than oriented towards any particular learning task. In this work, we present a formal model of DD, arguing that a precise characterization of the underlying optimization problem must specify the inference task associated with the application of interest. Without this task-specific focus, the DD problem is under-specified, and the selection of a DD algorithm for a particular task is merely heuristic. Our formalization reveals novel applications of DD across different modeling environments. We analyze existing DD methods through this broader lens, highlighting their strengths and limitations in terms of accuracy and faithfulness to optimal DD operation. Finally, we present numerical results for two case studies important in contemporary settings. Firstly, we address a critical challenge in medical data analysis: merging the knowledge from different datasets composed of intersecting, but not identical, sets of features, in order to construct a larger dataset in what is usually a small sample setting. Secondly, we consider out-of-distribution error across boundary conditions for physics-informed neural networks (PINNs), showing the potential for DD to provide more physically faithful data. By establishing this general formulation of DD, we aim to establish a new research paradigm by which DD can be understood and from which new DD techniques can arise.

Jianyang Gu, Haonan Wang, Ruoxi Jia et al.

Dataset distillation (DD) has witnessed significant progress in creating small datasets that encapsulate rich information from large original ones. Particularly, methods based on generative priors show promising performance, while maintaining computational efficiency and cross-architecture generalization. However, the generation process lacks explicit controllability for each sample. Previous distillation methods primarily match the real distribution from the perspective of the entire dataset, whereas overlooking concept completeness at the instance level. The missing or incorrectly represented object details cannot be efficiently compensated due to the constrained sample amount typical in DD settings. To this end, we propose incorporating the concept understanding of large language models (LLMs) to perform Concept-Informed Diffusion (CONCORD) for dataset distillation. Specifically, distinguishable and fine-grained concepts are retrieved based on category labels to inform the denoising process and refine essential object details. By integrating these concepts, the proposed method significantly enhances both the controllability and interpretability of the distilled image generation, without relying on pre-trained classifiers. We demonstrate the efficacy of CONCORD by achieving state-of-the-art performance on ImageNet-1K and its subsets. The code implementation is released in https://github.com/vimar-gu/CONCORD.

Qinsi Wang, Saeed Vahidian, Hancheng Ye et al.

Large language models (LLMs) with billions of parameters have sparked a new wave of exciting AI applications. However, their high computational costs and memory demands during inference pose significant challenges. Adaptive sparse activation inference, which activates only a small number of neurons for each token, offers a novel way to accelerate model inference without degrading performance, showing great potential for resource-constrained hardware devices. Nevertheless, existing methods predict activated neurons based on individual tokens with additional MLP, which involve frequent changes in activation maps and resource calls, limiting the acceleration benefits of sparse activation. In this paper, we introduce CoreInfer, an MLP-free adaptive sparse activation inference method based on sentence-level prediction. Specifically, we propose the concept of sentence-wise core neurons, which refers to the subset of neurons most critical for a given sentence, and empirically demonstrate its effectiveness. To determine the core neurons, we explore the correlation between core neurons and the sentence's semantics. Remarkably, we discovered that core neurons exhibit both stability and similarity in relation to the sentence's semantics -- an insight overlooked by previous studies. Building on this finding, we further design two semantic-based methods for predicting core neurons to fit different input scenarios. In CoreInfer, the core neurons are determined during the pre-filling stage and fixed during the encoding stage, enabling zero-cost sparse inference. We evaluated the model generalization and task generalization of CoreInfer across various models and tasks. Notably, on an NVIDIA TITAN XP GPU, CoreInfer achieved a 10.33 times and 2.72 times speedup compared to the Huggingface implementation and PowerInfer, respectively.

Jianyi Zhang, Saeed Vahidian, Martin Kuo et al.

While "instruction-tuned" generative large language models (LLMs) have demonstrated an impressive ability to generalize to new tasks, the training phases heavily rely on large amounts of diverse and high-quality instruction data (such as ChatGPT and GPT-4). Unfortunately, acquiring high-quality data, especially when it comes to human-written data, can pose significant challenges both in terms of cost and accessibility. Moreover, concerns related to privacy can further limit access to such data, making the process of obtaining it a complex and nuanced undertaking. Consequently, this hinders the generality of the tuned models and may restrict their effectiveness in certain contexts. To tackle this issue, our study introduces a new approach called Federated Instruction Tuning (FedIT), which leverages federated learning (FL) as the learning framework for the instruction tuning of LLMs. This marks the first exploration of FL-based instruction tuning for LLMs. This is especially important since text data is predominantly generated by end users. Therefore, it is imperative to design and adapt FL approaches to effectively leverage these users' diverse instructions stored on local devices, while preserving privacy and ensuring data security. In the current paper, by conducting widely used GPT-4 auto-evaluation, we demonstrate that by exploiting the heterogeneous and diverse sets of instructions on the client's end with the proposed framework FedIT, we improved the performance of LLMs compared to centralized training with only limited local instructions. Further, in this paper, we developed a Github repository named Shepherd. This repository offers a foundational framework for exploring federated fine-tuning of LLMs using heterogeneous instructions across diverse categories.

Saeed Vahidian, Mahdi Morafah, Bill Lin

The traditional approach in FL tries to learn a single global model collaboratively with the help of many clients under the orchestration of a central server. However, learning a single global model might not work well for all clients participating in the FL under data heterogeneity. Therefore, the personalization of the global model becomes crucial in handling the challenges that arise with statistical heterogeneity and the non-IID distribution of data. Unlike prior works, in this work we propose a new approach for obtaining a personalized model from a client-level objective. This further motivates all clients to participate in federation even under statistical heterogeneity in order to improve their performance, instead of merely being a source of data and model training for the central server. To realize this personalization, we leverage finding a small subnetwork for each client by applying hybrid pruning (combination of structured and unstructured pruning), and unstructured pruning. Through a range of experiments on different benchmarks, we observed that the clients with similar data (labels) share similar personal parameters. By finding a subnetwork for each client ...

Siavash Khodadadeh, Sharare Zehtabian, Saeed Vahidian et al.

Unsupervised meta-learning approaches rely on synthetic meta-tasks that are created using techniques such as random selection, clustering and/or augmentation. Unfortunately, clustering and augmentation are domain-dependent, and thus they require either manual tweaking or expensive learning. In this work, we describe an approach that generates meta-tasks using generative models. A critical component is a novel approach of sampling from the latent space that generates objects grouped into synthetic classes forming the training and validation data of a meta-task. We find that the proposed approach, LAtent Space Interpolation Unsupervised Meta-learning (LASIUM), outperforms or is competitive with current unsupervised learning baselines on few-shot classification tasks on the most widely used benchmark datasets. In addition, the approach promises to be applicable without manual tweaking over a wider range of domains than previous approaches.

Saeed Vahidian, Baharan Mirzasoleiman, Alexander Cloninger

In a number of situations, collecting a function value for every data point may be prohibitively expensive, and random sampling ignores any structure in the underlying data. We introduce a scalable optimization algorithm with no correction steps (in contrast to Frank-Wolfe and its variants), a variant of gradient ascent for coreset selection in graphs, that greedily selects a weighted subset of vertices that are deemed most important to sample. Our algorithm estimates the mean of the function by taking a weighted sum only at these vertices, and we provably bound the estimation error in terms of the location and weights of the selected vertices in the graph. In addition, we consider the case where nodes have different selection costs and provide bounds on the quality of the low-cost selected coresets. We demonstrate the benefits of our algorithm on the semi-supervised node classification of graph convolutional neural network, point clouds and structured graphs, as well as sensor placement where the cost of placing sensors depends on the location of the placement. We also elucidate that the empirical convergence of our proposed method is faster than random selection and various clustering methods while still respecting sensor placement cost. The paper concludes with validation of the developed algorithm on both synthetic and real datasets, demonstrating that it outperforms the current state of the art.

Saeed Vahidian, Sajad Hatamnia, Benoit Champagne

Physical layer security offers an efficient means to decrease the risk of confidential information leakage through wiretap links. In this paper, we address the physical-layer security in a cooperative wireless subnetwork that includes a source-destination pair and multiple relays, exchanging information in the presence of a malevolent eavesdropper. Specifically, the eavesdropper is active in the network and transmits artificial noise (AN) with a multiple-antenna transmitter to confound both the relays and the destination. We first analyse the secrecy capacity of the direct source-to-destination transmission in terms of intercept probability (IP) and secrecy outage probability (SOP). A decode-and-forward incremental relaying (IR) protocol is then introduced to improve reliability and security of communications in the presence of the active eavesdropper. Within this context, and depending on the availability of channel state information, three different schemes (one optimal and two sub-optimal) are proposed to select a trusted relay to improve the achievable secrecy rate. For each one of these schemes, and for both selection and maximum ratio combining at the destination and eavesdropper, we derive new and exact closed-form expressions for the IP and SOP. Our analysis and simulation results demonstrate the superior performance of the proposed IR-based selection schemes for secure communication. They also confirm the existence of a floor phenomenon for the SOP in the absence of AN.

Yasaman Ettefagh, Mohammad Hossein Moghaddam, Saeed Vahidian

In this research work, a novel framework is pro- posed as an efficient successor to traditional imaging methods for breast cancer detection in order to decrease the computational complexity. In this framework, the breast is devided into seg- ments in an iterative process and in each iteration, the one having the most probability of containing tumor with lowest possible resolution is selected by using suitable decision metrics. After finding the smallest tumor-containing segment, the resolution is increased in the detected tumor-containing segment, leaving the other parts of the breast image with low resolution. Our framework is applied on the most common used beamforming techniques, such as delay and sum (DAS) and delay multiply and sum (DMAS) and according to simulation results, our framework can decrease the computational complexity significantly for both DAS and DMAS without imposing any degradation on accuracy of basic algorithms. The amount of complexity reduction can be determined manually or automatically based on two proposed methods that are described in this framework.

Mohammad Hossein Moghaddam, Mohammad Javad Azizipour, Saeed Vahidian et al.

This paper introduces a framework for super-resolution of scalable video based on compressive sensing and sparse representation of residual frames in reconnaissance and surveillance applications. We exploit efficient compressive sampling and sparse reconstruction algorithms to super-resolve the video sequence with respect to different compression rates. We use the sparsity of residual information in residual frames as the key point in devising our framework. Moreover, a controlling factor as the compressibility threshold to control the complexity-performance trade-off is defined. Numerical experiments confirm the efficiency of the proposed framework in terms of the compression rate as well as the quality of reconstructed video sequence in terms of PSNR measure. The framework leads to a more efficient compression rate and higher video quality compared to other state-of-the-art algorithms considering performance-complexity trade-offs.