Existing AI-based point cloud compression methods struggle with dependence on specific training data distributions, which limits their real-world deployment. Implicit Neural Representation (INR) methods solve the above problem by encoding overfitted network parameters to the bitstream, resulting in more distribution-agnostic results. However, due to the limitation of encoding time and decoder size, current INR based methods only consider lossy geometry compression. In this paper, we propose the first INR based lossless point cloud geometry compression method called Lossless Implicit Neural Representations for Point Cloud Geometry Compression (LINR-PCGC). To accelerate encoding speed, we design a group of point clouds level coding framework with an effective network initialization strategy, which can reduce around 60% encoding time. A lightweight coding network based on multiscale SparseConv, consisting of scale context extraction, child node prediction, and model compression modules, is proposed to realize fast inference and compact decoder size. Experimental results show that our method consistently outperforms traditional and AI-based methods: for example, with the convergence time in the MVUB dataset, our method reduces the bitstream by approximately 21.21% compared to G-PCC TMC13v23 and 21.95% compared to SparsePCGC. Our project can be seen on https://huangwenjie2023.github.io/LINR-PCGC/.
Vision-Language-Action models (VLA) have demonstrated remarkable capabilities and promising potential in solving complex robotic manipulation tasks. However, their substantial parameter sizes and high inference latency pose significant challenges for real-world deployment, particularly on resource-constrained robotic platforms. To address this issue, we begin by conducting an extensive empirical study to explore the effectiveness of model compression techniques when applied to VLAs. Building on the insights gained from these preliminary experiments, we propose RLRC, a three-stage recovery method for compressed VLAs, including structured pruning, performance recovery based on SFT and RL, and further quantization. RLRC achieves up to an 8x reduction in memory usage and a 2.3x improvement in inference throughput, while maintaining or even surpassing the original VLA's task success rate. Extensive experiments show that RLRC consistently outperforms existing compression baselines, demonstrating strong potential for on-device deployment of VLAs. Project website: https://rlrc-vla.github.io
While modern segmentation models often prioritize performance over practicality, we advocate a design philosophy prioritizing simplicity and efficiency, and attempted high performance segmentation model design. This paper presents SimpleUNet, a scalable ultra-lightweight medical image segmentation model with three key innovations: (1) A partial feature selection mechanism in skip connections for redundancy reduction while enhancing segmentation performance; (2) A fixed-width architecture that prevents exponential parameter growth across network stages; (3) An adaptive feature fusion module achieving enhanced representation with minimal computational overhead. With a record-breaking 16 KB parameter configuration, SimpleUNet outperforms LBUNet and other lightweight benchmarks across multiple public datasets. The 0.67 MB variant achieves superior efficiency (8.60 GFLOPs) and accuracy, attaining a mean DSC/IoU of 85.76%/75.60% on multi-center breast lesion datasets, surpassing both U-Net and TransUNet. Evaluations on skin lesion datasets (ISIC 2017/2018: mDice 84.86%/88.77%) and endoscopic polyp segmentation (KVASIR-SEG: 86.46%/76.48% mDice/mIoU) confirm consistent dominance over state-of-the-art models. This work demonstrates that extreme model compression need not compromise performance, providing new insights for efficient and accurate medical image segmentation. Codes can be found at https://github.com/Frankyu5666666/SimpleUNet.
Large Language Models (LLMs) are central to many contemporary AI applications, yet their extensive parameter counts pose significant challenges for deployment in memory- and compute-constrained environments. Recent works in eXplainable AI (XAI), particularly on attribution methods, suggest that interpretability can also enable model compression by identifying and removing components irrelevant to inference. In this paper, we leverage Layer-wise Relevance Propagation (LRP) to perform attribution-guided pruning of LLMs. While LRP has shown promise in structured pruning for vision models, we extend it to unstructured pruning in LLMs and demonstrate that it can substantially reduce model size with minimal performance loss. Our method is especially effective in extracting task-relevant subgraphs -- so-called ``circuits'' -- which can represent core functions (e.g., indirect object identification). Building on this, we introduce a technique for model correction, by selectively removing circuits responsible for spurious behaviors (e.g., toxic outputs). All in all, we gather these techniques as a uniform holistic framework and showcase its effectiveness and limitations through extensive experiments for compression, circuit discovery and model correction on Llama and OPT models, highlighting its potential for improving both model efficiency and safety. Our code is publicly available at https://github.com/erfanhatefi/SparC3.
The ever-increasing computational demands and deployment costs of large language models (LLMs) have spurred numerous compressing methods. Compared to quantization and unstructured pruning, SVD compression offers superior hardware compatibility and theoretical guarantees. However, existing SVD-based methods focus on the overall discrepancy between the original and compressed matrices while overlooking the protection of critical components within the matrix, which leads to inferior performance in the compressed models. This paper proposes a dual-level importance protection mechanism to enhance SVD-based compression methods: (1) local importance protection: preserving the most critical singular vectors within each weight matrix through channel-weighted data whitening; and (2) global importance protection: enabling less important layers to bear a greater portion of the compression burden through either a heuristic or optimization-based approach, thereby minimizing the impact of compression on critical layers. Extensive experiments demonstrate that DipSVD outperforms existing SVD-based compression approaches across multiple benchmarks, achieving superior model performance especially at high model compression ratios.
This thesis addresses the technical challenges of applying machine learning to understand and interpret medical audio signals. The sounds of our lungs, heart, and voice convey vital information about our health. Yet, in contemporary medicine, these sounds are primarily analyzed through auditory interpretation by experts using devices like stethoscopes. Automated analysis offers the potential to standardize the processing of medical sounds, enable screening in low-resource settings where physicians are scarce, and detect subtle patterns that may elude human perception, thereby facilitating early diagnosis and treatment. Focusing on the analysis of infant cry sounds to predict medical conditions, this thesis contributes on four key fronts. First, in low-data settings, we demonstrate that large databases of adult speech can be harnessed through neural transfer learning to develop more accurate and robust models for infant cry analysis. Second, in cost-effective modeling, we introduce an end-to-end model compression approach for recurrent networks using tensor decomposition. Our method requires no post-hoc processing, achieves compression rates of several hundred-fold, and delivers accurate, portable models suitable for resource-constrained devices. Third, we propose novel domain adaptation techniques tailored for audio models and adapt existing methods from computer vision. These approaches address dataset bias and enhance generalization across domains while maintaining strong performance on the original data. Finally, to advance research in this domain, we release a unique, open-source dataset of infant cry sounds, developed in collaboration with clinicians worldwide. This work lays the foundation for recognizing the infant cry as a vital sign and highlights the transformative potential of AI-driven audio monitoring in shaping the future of accessible and affordable healthcare.
Model compression is critical for deploying deep learning models on resource-constrained devices. We introduce a novel method enhancing knowledge distillation with integrated gradients (IG) as a data augmentation strategy. Our approach overlays IG maps onto input images during training, providing student models with deeper insights into teacher models' decision-making processes. Extensive evaluation on CIFAR-10 demonstrates that our IG-augmented knowledge distillation achieves 92.6% testing accuracy with a 4.1x compression factor-a significant 1.1 percentage point improvement ($p<0.001$) over non-distilled models (91.5%). This compression reduces inference time from 140 ms to 13 ms. Our method precomputes IG maps before training, transforming substantial runtime costs into a one-time preprocessing step. Our comprehensive experiments include: (1) comparisons with attention transfer, revealing complementary benefits when combined with our approach; (2) Monte Carlo simulations confirming statistical robustness; (3) systematic evaluation of compression factor versus accuracy trade-offs across a wide range (2.2x-1122x); and (4) validation on an ImageNet subset aligned with CIFAR-10 classes, demonstrating generalisability beyond the initial dataset. These extensive ablation studies confirm that IG-based knowledge distillation consistently outperforms conventional approaches across varied architectures and compression ratios. Our results establish this framework as a viable compression technique for real-world deployment on edge devices while maintaining competitive accuracy.
Small-Footprint Keyword Spotting (SF-KWS) has gained popularity in today's landscape of smart voice-activated devices, smartphones, and Internet of Things (IoT) applications. This surge is attributed to the advancements in Deep Learning, enabling the identification of predefined words or keywords from a continuous stream of words. To implement the SF-KWS model on edge devices with low power and limited memory in real-world scenarios, a efficient Tiny Machine Learning (TinyML) framework is essential. In this study, we explore seven distinct categories of techniques namely, Model Architecture, Learning Techniques, Model Compression, Attention Awareness Architecture, Feature Optimization, Neural Network Search, and Hybrid Approaches, which are suitable for developing an SF-KWS system. This comprehensive overview will serve as a valuable resource for those looking to understand, utilize, or contribute to the field of SF-KWS. The analysis conducted in this work enables the identification of numerous potential research directions, encompassing insights from automatic speech recognition research and those specifically pertinent to the realm of spoken SF-KWS.
Mixture-of-Experts (MoE) models have emerged as a cornerstone of large-scale deep learning by efficiently distributing computation and enhancing performance. However, their unique architecture-characterized by sparse expert activation and dynamic routing mechanisms-introduces inherent complexities that challenge conventional quantization techniques. Existing post-training quantization (PTQ) methods struggle to address activation outliers, router consistency and sparse expert calibration, leading to significant performance degradation. To bridge this gap, we propose EAQuant, a novel PTQ framework tailored for MoE architectures. Our method systematically tackles these challenges through three key innovations: (1) expert-aware smoothing aggregation to suppress activation outliers and stabilize quantization, (2) router logits distribution alignment to preserve expert selection consistency post-quantization, and (3) expert-level calibration data balance to optimize sparsely activated experts. Extensive experiments across W4A4 and extreme W3A4 quantization configurations demonstrate that EAQuant significantly outperforms existing methods, achieving average score improvements of 1.15 - 2.28% across three diverse MoE architectures, with particularly pronounced gains in reasoning tasks and robust performance retention under aggressive quantization. By integrating these innovations, EAQuant establishes a new state-of-the-art for high-precision, efficient MoE model compression. Our code is available at https://github.com/darren-fzq/EAQuant.
Post-training quantization (PTQ) has stood out as a cost-effective and promising model compression paradigm in recent years, as it avoids computationally intensive model retraining. Nevertheless, current PTQ methods for Vision Transformers (ViTs) still suffer from significant accuracy degradation, especially under low-bit quantization. To address these shortcomings, we analyze the prevailing Hessian-guided quantization loss, and uncover certain limitations of conventional Hessian approximations. By following the block-wise reconstruction framework, we propose a novel PTQ method for ViTs, dubbed FIMA-Q. Specifically, we firstly establish the connection between KL divergence and FIM, which enables fast computation of the quantization loss during reconstruction. We further propose an efficient FIM approximation method, namely DPLR-FIM, by employing the diagonal plus low-rank principle, and formulate the ultimate quantization loss. Our extensive experiments, conducted across various vision tasks with representative ViT-based architectures on public datasets, demonstrate that our method substantially promotes the accuracy compared to the state-of-the-art approaches, especially in the case of low-bit quantization. The source code is available at https://github.com/ShiheWang/FIMA-Q.
Spiking Neural Networks (SNNs), known for their potential to enable low energy consumption and computational cost, can bring significant advantages to the realm of embedded machine learning for edge applications. However, input coming from standard digital sensors must be encoded into spike trains before it can be elaborated with neuromorphic computing technologies. We present here a detailed comparison of available spike encoding techniques for the translation of time-varying signals into the event-based signal domain, tested on two different datasets both acquired through commercially available digital devices: the Free Spoken Digit dataset (FSD), consisting of 8-kHz audio files, and the WISDM dataset, composed of 20-Hz recordings of human activity through mobile and wearable inertial sensors. We propose a complete pipeline to benchmark these encoding techniques by performing time-dependent signal classification through a Spiking Convolutional Neural Network (sCNN), including a signal preprocessing step consisting of a bank of filters inspired by the human cochlea, feature extraction by production of a sonogram, transfer learning via an equivalent ANN, and model compression schemes aimed at resource optimization. The resulting performance comparison and analysis provides a powerful practical tool, empowering developers to select the most suitable coding method based on the type of data and the desired processing algorithms, and further expands the applicability of neuromorphic computational paradigms to embedded sensor systems widely employed in the IoT and industrial domains.
The high computational costs associated with large deep learning models significantly hinder their practical deployment. Model pruning has been widely explored in deep learning literature to reduce their computational burden, but its application has been largely limited to computer vision tasks such as image classification and object detection. In this work, we propose a structured pruning method targeted for Learned Image Compression (LIC) models that aims to reduce the computational costs associated with image compression while maintaining the rate-distortion performance. We employ a Neural Architecture Search (NAS) method based on the rate-distortion loss for computing the pruning ratio for each layer of the network. We compare our pruned model with the uncompressed LIC Model with same network architecture and show that it can achieve model size reduction without any BD-Rate performance drop. We further show that our pruning method can be integrated with model quantization to achieve further model compression while maintaining similar BD-Rate performance. We have made the source code available at gitlab.com/viper-purdue/lic-pruning.
Video matting is crucial for applications such as film production and virtual reality, yet deploying its computationally intensive models on resource-constrained devices presents challenges. Quantization is a key technique for model compression and acceleration. As an efficient approach, Post-Training Quantization (PTQ) is still in its nascent stages for video matting, facing significant hurdles in maintaining accuracy and temporal coherence. To address these challenges, this paper proposes a novel and general PTQ framework specifically designed for video matting models, marking, to the best of our knowledge, the first systematic attempt in this domain. Our contributions include: (1) A two-stage PTQ strategy that combines block-reconstruction-based optimization for fast, stable initial quantization and local dependency capture, followed by a global calibration of quantization parameters to minimize accuracy loss. (2) A Statistically-Driven Global Affine Calibration (GAC) method that enables the network to compensate for cumulative statistical distortions arising from factors such as neglected BN layer effects, even reducing the error of existing PTQ methods on video matting tasks up to 20%. (3) An Optical Flow Assistance (OFA) component that leverages temporal and semantic priors from frames to guide the PTQ process, enhancing the model's ability to distinguish moving foregrounds in complex scenes and ultimately achieving near full-precision performance even under ultra-low-bit quantization. Comprehensive quantitative and visual results show that our PTQ4VM achieves the state-of-the-art accuracy performance across different bit-widths compared to the existing quantization methods. We highlight that the 4-bit PTQ4VM even achieves performance close to the full-precision counterpart while enjoying 8x FLOP savings.
To address the enormous size of Large Language Models (LLMs), model compression methods, such as quantization and pruning, are often deployed, especially on edge devices. In this work, we focus on layer-wise post-training quantization and pruning. Drawing connections between activation-aware weight pruning and sparse approximation problems, and motivated by the success of Iterative Hard Thresholding (IHT), we propose a unified method for Activation-aware Weight pruning and quantization via Projected gradient descent (AWP). Our experiments demonstrate that AWP outperforms state-of-the-art LLM pruning and quantization methods. Theoretical convergence guarantees of the proposed method for pruning are also provided.
Efficiently serving large language models (LLMs) under dynamic and bursty workloads remains a key challenge for real-world deployment. Existing serving frameworks and static model compression techniques fail to adapt to workload fluctuations, leading to either service-level objective (SLO) violations under full-precision serving or persistent accuracy degradation with static quantization. We present MorphServe, a dynamic, workload-aware LLM serving framework based on morphological adaptation. MorphServe introduces two asynchronous, token-level runtime mechanisms: quantized layer swapping, which selectively replaces less impactful layers with quantized alternatives during high-load periods, and pressure-aware KV cache resizing, which dynamically adjusts KV cache capacity in response to memory pressure. These mechanisms enable state-preserving transitions with minimum runtime overhead and are fully compatible with modern scheduling and attention techniques. Extensive experiments on Vicuna and Llama family models with real-world workloads demonstrate that MorphServe reduces average SLO violations by 92.45 percent and improves the P95 TTFT latency by 2.2x-3.9x compared to full-precision serving, without compromising generation quality. These results establish MorphServe as a practical and elastic solution for LLM deployment in dynamic environments.
Large Language Models (LLMs) have enabled remarkable progress in natural language processing, yet their high computational and memory demands pose challenges for deployment in resource-constrained environments. Although recent low-rank decomposition methods offer a promising path for structural compression, they often suffer from accuracy degradation, expensive calibration procedures, and result in inefficient model architectures that hinder real-world inference speedups. In this paper, we propose FLAT-LLM, a fast and accurate, training-free structural compression method based on fine-grained low-rank transformations in the activation space. Specifically, we reduce the hidden dimension by transforming the weights using truncated eigenvectors computed via head-wise Principal Component Analysis (PCA), and employ an importance-based metric to adaptively allocate ranks across decoders. FLAT-LLM achieves efficient and effective weight compression without recovery fine-tuning, which could complete the calibration within a few minutes. Evaluated across 4 models and 11 datasets, FLAT-LLM outperforms structural pruning baselines in generalization and downstream performance, while delivering inference speedups over decomposition-based methods.
Compressing and pruning large machine learning models has become a critical step towards their deployment in real-world applications. Standard pruning and compression techniques are typically designed without taking the structure of the network's weights into account, limiting their effectiveness. We explore the impact of smooth regularization on neural network training and model compression. By applying nuclear norm, first- and second-order derivative penalties of the weights during training, we encourage structured smoothness while preserving predictive performance on par with non-smooth models. We find that standard pruning methods often perform better when applied to these smooth models. Building on this observation, we apply a Singular-Value-Decomposition-based compression method that exploits the underlying smooth structure and approximates the model's weight tensors by smaller low-rank tensors. Our approach enables state-of-the-art compression without any fine-tuning - reaching up to $91\%$ accuracy on a smooth ResNet-18 on CIFAR-10 with $70\%$ fewer parameters.
In-network computation represents a transformative approach to addressing the escalating demands of Artificial Intelligence (AI) workloads on network infrastructure. By leveraging the processing capabilities of network devices such as switches, routers, and Network Interface Cards (NICs), this paradigm enables AI computations to be performed directly within the network fabric, significantly reducing latency, enhancing throughput, and optimizing resource utilization. This paper provides a comprehensive analysis of optimizing in-network computation for AI, exploring the evolution of programmable network architectures, such as Software-Defined Networking (SDN) and Programmable Data Planes (PDPs), and their convergence with AI. It examines methodologies for mapping AI models onto resource-constrained network devices, addressing challenges like limited memory and computational capabilities through efficient algorithm design and model compression techniques. The paper also highlights advancements in distributed learning, particularly in-network aggregation, and the potential of federated learning to enhance privacy and scalability. Frameworks like Planter and Quark are discussed for simplifying development, alongside key applications such as intelligent network monitoring, intrusion detection, traffic management, and Edge AI. Future research directions, including runtime programmability, standardized benchmarks, and new applications paradigms, are proposed to advance this rapidly evolving field. This survey underscores the potential of in-network AI to create intelligent, efficient, and responsive networks capable of meeting the demands of next-generation AI applications.
This paper presents a novel approach for speech foundation models compression that tightly integrates model pruning and parameter update into a single stage. Highly compact layer-level tied self-pinching gates each containing only a single learnable threshold are jointly trained with uncompressed models and used in fine-grained neuron level pruning. Experiments conducted on the LibriSpeech-100hr corpus suggest that our approach reduces the number of parameters of wav2vec2.0-base and HuBERT-large models by 65% and 60% respectively, while incurring no statistically significant word error rate (WER) increase on the test-clean dataset. Compared to previously published methods on the same task, our approach not only achieves the lowest WER of 7.05% on the test-clean dataset under a comparable model compression ratio of 4.26x, but also operates with at least 25% less model compression time.
Tensorizing a neural network involves reshaping some or all of its dense weight matrices into higher-order tensors and approximating them using low-rank tensor network decompositions. This technique has shown promise as a model compression strategy for large-scale neural networks. However, despite encouraging empirical results, tensorized neural networks (TNNs) remain underutilized in mainstream deep learning. In this position paper, we offer a perspective on both the potential and current limitations of TNNs. We argue that TNNs represent a powerful yet underexplored framework for deep learning--one that deserves greater attention from both engineering and theoretical communities. Beyond compression, we highlight the value of TNNs as a flexible class of architectures with distinctive scaling properties and increased interpretability. A central feature of TNNs is the presence of bond indices, which introduce new latent spaces not found in conventional networks. These internal representations may provide deeper insight into the evolution of features across layers, potentially advancing the goals of mechanistic interpretability. We conclude by outlining several key research directions aimed at overcoming the practical barriers to scaling and adopting TNNs in modern deep learning workflows.