arXiv subjects

Yi Ma

Publications and source records attributed to Yi Ma.

At least 127 records · Page 7Linked to original sources

arXiv2024

White-Box Transformers via Sparse Rate Reduction: Compression Is All There Is?

In this paper, we contend that a natural objective of representation learning is to compress and transform the distribution of the data, say sets of tokens, towards a low-dimensional Gaussian mixture supported on incoherent subspaces. The goodness of such a representation can be evaluated by a principled measure, called sparse rate reduction, that simultaneously maximizes the intrinsic information gain and extrinsic sparsity of the learned representation. From this perspective, popular deep network architectures, including transformers, can be viewed as realizing iterative schemes to optimize this measure. Particularly, we derive a transformer block from alternating optimization on parts of this objective: the multi-head self-attention operator compresses the representation by implementing an approximate gradient descent step on the coding rate of the features, and the subsequent multi-layer perceptron sparsifies the features. This leads to a family of white-box transformer-like deep network architectures, named CRATE, which are mathematically fully interpretable. We show, by way of a novel connection between denoising and compression, that the inverse to the aforementioned compressive encoding can be realized by the same class of CRATE architectures. Thus, the so-derived white-box architectures are universal to both encoders and decoders. Experiments show that these networks, despite their simplicity, indeed learn to compress and sparsify representations of large-scale real-world image and text datasets, and achieve performance very close to highly engineered transformer-based models: ViT, MAE, DINO, BERT, and GPT2. We believe the proposed computational framework demonstrates great potential in bridging the gap between theory and practice of deep learning, from a unified perspective of data compression. Code is available at: https://ma-lab-berkeley.github.io/CRATE .

cs.LG↗

arXiv2024

Fast Iterative ELAA-MIMO Detection Exploiting Static Channel Components

Extremely large aperture array (ELAA) is a promising multiple-input multiple-output (MIMO) technique for next generation mobile networks. In this paper, we propose two novel approaches to accelerate the convergence of current iterative MIMO detectors in ELAA channels. Our approaches exploit the static components of the ELAA channel, which include line of sight (LoS) paths and deterministic non-LoS (NLoS) components due to channel hardening effects. This paper proposes novel convergence acceleration techniques for fast iterative ELAA-MIMO detection by leveraging the static channel component, including the LoS paths and deterministic NLoS components that arise due to channel hardening. Specifically, these static channel components are utilized in two ways: as preconditioning matrices for general iterative algorithms, and as initialization for quasi-Newton (QN) methods. Simulation results show that the proposed approaches converge significantly faster compared to current iterative MIMO detectors, especially under strong LoS conditions with high Rician K-factor. Furthermore, QN methods with the proposed initialization matrix consistently achieve the best convergence performance while maintaining low complexity.

cs.IT↗

arXiv2024

Accelerating Iteratively Linear Detectors in Multi-User (ELAA-)MIMO Systems with UW-SVD

Current iterative multiple-input multiple-output (MIMO) detectors suffer from slow convergence when the wireless channel is ill-conditioned. The ill-conditioning is mainly caused by spatial correlation between channel columns corresponding to the same user equipment, known as intra-user interference. In addition, in the emerging MIMO systems using an extremely large aperture array (ELAA), spatial non-stationarity can make the channel even more ill-conditioned. In this paper, user-wise singular value decomposition (UW-SVD) is proposed to accelerate the convergence of iterative MIMO detectors. Its basic principle is to perform SVD on each user's sub-channel matrix to eliminate intra-user interference. Then, the MIMO signal model is effectively transformed into an equivalent signal (e-signal) model, comprising an e-channel matrix and an e-signal vector. Existing iterative algorithms can be used to recover the e-signal vector, which undergoes post-processing to obtain the signal vector. It is proven that the e-channel matrix is better conditioned than the original MIMO channel for spatially correlated (ELAA-)MIMO channels. This implies that UW-SVD can accelerate current iterative algorithms, which is confirmed by our simulation results. Specifically, it can speed up convergence by up to 10 times in both uncoded and coded systems.

cs.IT↗

arXiv2024

ENOTO: Improving Offline-to-Online Reinforcement Learning with Q-Ensembles

Offline reinforcement learning (RL) is a learning paradigm where an agent learns from a fixed dataset of experience. However, learning solely from a static dataset can limit the performance due to the lack of exploration. To overcome it, offline-to-online RL combines offline pre-training with online fine-tuning, which enables the agent to further refine its policy by interacting with the environment in real-time. Despite its benefits, existing offline-to-online RL methods suffer from performance degradation and slow improvement during the online phase. To tackle these challenges, we propose a novel framework called ENsemble-based Offline-To-Online (ENOTO) RL. By increasing the number of Q-networks, we seamlessly bridge offline pre-training and online fine-tuning without degrading performance. Moreover, to expedite online performance enhancement, we appropriately loosen the pessimism of Q-value estimation and incorporate ensemble-based exploration mechanisms into our framework. Experimental results demonstrate that ENOTO can substantially improve the training stability, learning efficiency, and final performance of existing offline RL methods during online fine-tuning on a range of locomotion and navigation tasks, significantly outperforming existing offline-to-online RL methods.

cs.LG↗

arXiv2024

Wi-Fi Beyond Communications: Experimental Evaluation of Respiration Monitoring and Motion Detection Using COTS Devices

Wi-Fi sensing has become an attractive option for non-invasive monitoring of human activities and vital signs. This paper explores the feasibility of using state-of-the-art commercial off-the-shelf (COTS) devices for Wi-Fi sensing applications, particularly respiration monitoring and motion detection. We utilize the Intel AX210 network interface card (NIC) to transmit Wi-Fi signals in both 2.4 GHz and 6 GHz frequency bands. Our experiments rely on channel frequency response (CFR) and received signal strength indicator (RSSI) data, which are processed using a moving average algorithm to extract human behavior patterns. The experimental results demonstrate the effectiveness of our approach in capturing and representing human respiration and motion patterns. Furthermore, we compare the performance of Wi-Fi sensing across different frequency bands, highlighting the advantages of using higher frequencies for improved sensitivity and clarity. Our findings showcase the practicality of using COTS devices for Wi-Fi sensing and lay the groundwork for the development of non-invasive, contactless sensing systems. These systems have potential applications in various fields, including healthcare, smart homes, and Metaverse.

cs.IT↗

arXiv2024

Multi-Satellite MIMO Systems for Direct User-Satellite Communications: A Survey

Advancements in satellite technology have made direct-to-device connectivity a viable solution for ensuring global access. This method is designed to provide internet connectivity to remote, rural, or underserved areas where traditional cellular or broadband networks are lacking or insufficient. This paper is a survey providing an in-depth review of multi-satellite Multiple Input Multiple Output (MIMO) systems as a potential solution for addressing the link budget challenge in direct user-satellite communication. Special attention is given to works considering multi-satellite MIMO systems, both with and without satellite collaboration. In this context, collaboration refers to sharing data between satellites to improve the performance of the system. This survey begins by explaining several fundamental aspects of satellite communications (SatComs), which are vital prerequisites before investigating the multi-satellite MIMO systems. These aspects encompass satellite orbits, the structure of satellite systems, SatCom links, including the inter-satellite links (ISL) which facilitate satellite cooperation, satellite frequency bands, satellite antenna design, and satellite channel models, which should be known or estimated for effective data transmission to and from multiple satellites. Furthermore, this survey distinguishes itself by providing more comprehensive insights in comparison to other surveys. It specifically delves into the Orthogonal Time Frequency Space (OTFS) within the channel model section. It goes into detail about ISL noise and channel models, and it extends the ISL section by thoroughly investigating hybrid FSO/RF ISLs. Furthermore, analytical comparisons of simulation results from these works are presented to highlight the advantages of employing multi-satellite MIMO systems.

eess.SP↗

arXiv2024

Ctrl123: Consistent Novel View Synthesis via Closed-Loop Transcription

Large image diffusion models have demonstrated zero-shot capability in novel view synthesis (NVS). However, existing diffusion-based NVS methods struggle to generate novel views that are accurately consistent with the corresponding ground truth poses and appearances, even on the training set. This consequently limits the performance of downstream tasks, such as image-to-multiview generation and 3D reconstruction. We realize that such inconsistency is largely due to the fact that it is difficult to enforce accurate pose and appearance alignment directly in the diffusion training, as mostly done by existing methods such as Zero123. To remedy this problem, we propose Ctrl123, a closed-loop transcription-based NVS diffusion method that enforces alignment between the generated view and ground truth in a pose-sensitive feature space. Our extensive experiments demonstrate the effectiveness of Ctrl123 on the tasks of NVS and 3D reconstruction, achieving significant improvements in both multiview-consistency and pose-consistency over existing methods.

cs.CV↗

arXiv2024

A Survey of Computation Offloading with Task Types

Computation task offloading plays a crucial role in facilitating computation-intensive applications and edge intelligence, particularly in response to the explosive growth of massive data generation. Various enabling techniques, wireless technologies and mechanisms have already been proposed for task offloading, primarily aimed at improving the quality of services (QoS) for users. While there exists an extensive body of literature on this topic, exploring computation offloading from the standpoint of task types has been relatively underrepresented. This motivates our survey, which seeks to classify the state-of-the-art (SoTA) from the task type point-of-view. To achieve this, a thorough literature review is conducted to reveal the SoTA from various aspects, including architecture, objective, offloading strategy, and task types, with the consideration of task generation. It has been observed that task types are associated with data and have an impact on the offloading process, including elements like resource allocation and task assignment. Building upon this insight, computation offloading is categorized into two groups based on task types: static task-based offloading and dynamic task-based offloading. Finally, a prospective view of the challenges and opportunities in the field of future computation offloading is presented.

cs.DC↗

arXiv2024

Convergence Acceleration in Wireless Federated Learning: A Stackelberg Game Approach

This paper studies issues that arise with respect to the joint optimization for convergence time in federated learning over wireless networks (FLOWN). We consider the criterion and protocol for selection of participating devices in FLOWN under the energy constraint and derive its impact on device selection. In order to improve the training efficiency, age-of-information (AoI) enables FLOWN to assess the freshness of gradient updates among participants. Aiming to speed up convergence, we jointly investigate global loss minimization and latency minimization in a Stackelberg game based framework. Specifically, we formulate global loss minimization as a leader-level problem for reducing the number of required rounds, and latency minimization as a follower-level problem to reduce time consumption of each round. By decoupling the follower-level problem into two sub-problems, including resource allocation and sub-channel assignment, we achieve an optimal strategy of the follower through monotonic optimization and matching theory. At the leader-level, we derive an upper bound of convergence rate and subsequently reformulate the global loss minimization problem and propose a new age-of-update (AoU) based device selection algorithm. Simulation results indicate the superior performance of the proposed AoU based device selection scheme in terms of the convergence rate, as well as efficient utilization of available sub-channels.

cs.LG↗

arXiv2024

How Do Neural Spoofing Countermeasures Detect Partially Spoofed Audio?

Partially manipulating a sentence can greatly change its meaning. Recent work shows that countermeasures (CMs) trained on partially spoofed audio can effectively detect such spoofing. However, the current understanding of the decision-making process of CMs is limited. We utilize Grad-CAM and introduce a quantitative analysis metric to interpret CMs' decisions. We find that CMs prioritize the artifacts of transition regions created when concatenating bona fide and spoofed audio. This focus differs from that of CMs trained on fully spoofed audio, which concentrate on the pattern differences between bona fide and spoofed parts. Our further investigation explains the varying nature of CMs' focus while making correct or incorrect predictions. These insights provide a basis for the design of CM models and the creation of datasets. Moreover, this work lays a foundation of interpretability in the field of partial spoofed audio detection that has not been well explored previously.

eess.AS↗

arXiv2024

Near and Far Field Model Mismatch: Implications on 6G Communications, Localization, and Sensing

The upcoming 6G technology is expected to operate in near-field (NF) radiating conditions thanks to high-frequency and electrically large antenna arrays. Although several studies have already addressed this possibility, it is worth noting that NF models introduce higher complexity, the justification for which is not always evident in terms of performance improvements. This article investigates the implications of the mismatch between NF and far-field (FF) models concerning communication, localization, and sensing systems. Such disparity can lead to a degradation of performance metrics such as sensing and localization accuracy and communication efficiency. By exploring the effects of mismatches between NF and FF models, this study seeks to revolve around the challenges faced by system designers, offering insights about the balance between model accuracy and achievable performance. Finally, we conduct a numerical performance analysis to verify the impact of the mismatch between NF and FF models.

eess.SP↗

arXiv2024

Image Clustering via the Principle of Rate Reduction in the Age of Pretrained Models

The advent of large pre-trained models has brought about a paradigm shift in both visual representation learning and natural language processing. However, clustering unlabeled images, as a fundamental and classic machine learning problem, still lacks an effective solution, particularly for large-scale datasets. In this paper, we propose a novel image clustering pipeline that leverages the powerful feature representation of large pre-trained models such as CLIP and cluster images effectively and efficiently at scale. We first developed a novel algorithm to estimate the number of clusters in a given dataset. We then show that the pre-trained features are significantly more structured by further optimizing the rate reduction objective. The resulting features may significantly improve the clustering accuracy, e.g., from 57\% to 66\% on ImageNet-1k. Furthermore, by leveraging CLIP's multimodality bridge between image and text, we develop a simple yet effective self-labeling algorithm that produces meaningful captions for the clusters. Through extensive experiments, we show that our pipeline works well on standard datasets such as CIFAR-10, CIFAR-100, and ImageNet-1k. It also extends to datasets that are not curated for clustering, such as LAION-Aesthetics and WikiArts. We released the code in https://github.com/LeslieTrue/CPP.

cs.CV↗

arXiv2024

Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs

Is vision good enough for language? Recent advancements in multimodal models primarily stem from the powerful reasoning abilities of large language models (LLMs). However, the visual component typically depends only on the instance-level contrastive language-image pre-training (CLIP). Our research reveals that the visual capabilities in recent multimodal LLMs (MLLMs) still exhibit systematic shortcomings. To understand the roots of these errors, we explore the gap between the visual embedding space of CLIP and vision-only self-supervised learning. We identify ''CLIP-blind pairs'' - images that CLIP perceives as similar despite their clear visual differences. With these pairs, we construct the Multimodal Visual Patterns (MMVP) benchmark. MMVP exposes areas where state-of-the-art systems, including GPT-4V, struggle with straightforward questions across nine basic visual patterns, often providing incorrect answers and hallucinated explanations. We further evaluate various CLIP-based vision-and-language models and found a notable correlation between visual patterns that challenge CLIP models and those problematic for multimodal LLMs. As an initial effort to address these issues, we propose a Mixture of Features (MoF) approach, demonstrating that integrating vision self-supervised learning features with MLLMs can significantly enhance their visual grounding capabilities. Together, our research suggests visual representation learning remains an open challenge, and accurate visual grounding is crucial for future successful multimodal systems.

cs.CV↗

arXiv2024

Strongly Gauduchon Hyperbolicity and Two Other Types of Hyperbolicity

This paper proposes sG-hyperbolicity as a new tool for studying hyperbolicity on complex manifolds. It demonstrates that this notion leads to a wider class of divisorially hyperbolic manifolds compared to balanced hyperbolicity. We also introduce weakly p-Kähler hyperbolic structures and pluriclosed star split hyperbolic metrics as possible new avenues for exploration.

math.DG↗

arXiv2024

Masked Completion via Structured Diffusion with White-Box Transformers

Modern learning frameworks often train deep neural networks with massive amounts of unlabeled data to learn representations by solving simple pretext tasks, then use the representations as foundations for downstream tasks. These networks are empirically designed; as such, they are usually not interpretable, their representations are not structured, and their designs are potentially redundant. White-box deep networks, in which each layer explicitly identifies and transforms structures in the data, present a promising alternative. However, existing white-box architectures have only been shown to work at scale in supervised settings with labeled data, such as classification. In this work, we provide the first instantiation of the white-box design paradigm that can be applied to large-scale unsupervised representation learning. We do this by exploiting a fundamental connection between diffusion, compression, and (masked) completion, deriving a deep transformer-like masked autoencoder architecture, called CRATE-MAE, in which the role of each layer is mathematically fully interpretable: they transform the data distribution to and from a structured representation. Extensive empirical evaluations confirm our analytical insights. CRATE-MAE demonstrates highly promising performance on large-scale imagery datasets while using only ~30% of the parameters compared to the standard masked autoencoder with the same model configuration. The representations learned by CRATE-MAE have explicit structure and also contain semantic meaning. Code is available at https://github.com/Ma-Lab-Berkeley/CRATE .

cs.LG↗

arXiv2024

Uni-RLHF: Universal Platform and Benchmark Suite for Reinforcement Learning with Diverse Human Feedback

Reinforcement Learning with Human Feedback (RLHF) has received significant attention for performing tasks without the need for costly manual reward design by aligning human preferences. It is crucial to consider diverse human feedback types and various learning methods in different environments. However, quantifying progress in RLHF with diverse feedback is challenging due to the lack of standardized annotation platforms and widely used unified benchmarks. To bridge this gap, we introduce Uni-RLHF, a comprehensive system implementation tailored for RLHF. It aims to provide a complete workflow from real human feedback, fostering progress in the development of practical problems. Uni-RLHF contains three packages: 1) a universal multi-feedback annotation platform, 2) large-scale crowdsourced feedback datasets, and 3) modular offline RLHF baseline implementations. Uni-RLHF develops a user-friendly annotation interface tailored to various feedback types, compatible with a wide range of mainstream RL environments. We then establish a systematic pipeline of crowdsourced annotations, resulting in large-scale annotated datasets comprising more than 15 million steps across 30+ popular tasks. Through extensive experiments, the results in the collected datasets demonstrate competitive performance compared to those from well-designed manual rewards. We evaluate various design choices and offer insights into their strengths and potential areas of improvement. We wish to build valuable open-source platforms, datasets, and baselines to facilitate the development of more robust and reliable RLHF solutions based on realistic human feedback. The website is available at https://uni-rlhf.github.io/.

cs.LG↗

arXiv2024

RLIF: Interactive Imitation Learning as Reinforcement Learning

Although reinforcement learning methods offer a powerful framework for automatic skill acquisition, for practical learning-based control problems in domains such as robotics, imitation learning often provides a more convenient and accessible alternative. In particular, an interactive imitation learning method such as DAgger, which queries a near-optimal expert to intervene online to collect correction data for addressing the distributional shift challenges that afflict naïve behavioral cloning, can enjoy good performance both in theory and practice without requiring manually specified reward functions and other components of full reinforcement learning methods. In this paper, we explore how off-policy reinforcement learning can enable improved performance under assumptions that are similar but potentially even more practical than those of interactive imitation learning. Our proposed method uses reinforcement learning with user intervention signals themselves as rewards. This relaxes the assumption that intervening experts in interactive imitation learning should be near-optimal and enables the algorithm to learn behaviors that improve over the potential suboptimal human expert. We also provide a unified framework to analyze our RL method and DAgger; for which we present the asymptotic analysis of the suboptimal gap for both methods as well as the non-asymptotic sample complexity bound of our method. We then evaluate our method on challenging high-dimensional continuous control simulation benchmarks as well as real-world robotic vision-based manipulation tasks. The results show that it strongly outperforms DAgger-like approaches across the different tasks, especially when the intervening experts are suboptimal. Code and videos can be found on the project website: https://rlif-page.github.io

cs.AI↗

arXiv2024

Bit Rate Matching Algorithm Optimization in JPEG-AI Verification Model

The research on neural network (NN) based image compression has shown superior performance compared to classical compression frameworks. Unlike the hand-engineered transforms in the classical frameworks, NN-based models learn the non-linear transforms providing more compact bit representations, and achieve faster coding speed on parallel devices over their classical counterparts. Those properties evoked the attention of both scientific and industrial communities, resulting in the standardization activity JPEG-AI. The verification model for the standardization process of JPEG-AI is already in development and has surpassed the advanced VVC intra codec. To generate reconstructed images with the desired bits per pixel and assess the BD-rate performance of both the JPEG-AI verification model and VVC intra, bit rate matching is employed. However, the current state of the JPEG-AI verification model experiences significant slowdowns during bit rate matching, resulting in suboptimal performance due to an unsuitable model. The proposed methodology offers a gradual algorithmic optimization for matching bit rates, resulting in a fourfold acceleration and over 1% improvement in BD-rate at the base operation point. At the high operation point, the acceleration increases up to sixfold.

cs.CV↗