Search Results for author:
Found 44 papers, 13 papers with code
EfficientFormer: Vision Transformers at MobileNet Speed
Our work proves that properly designed transformers can reach extremely low latency on mobile devices while maintaining high performance.
YOLObile: Real-Time Object Detection on Mobile Devices via Compression-Compilation Co-Design
In this work, we propose YOLObile framework, a real-time object detection on mobile devices via compression-compilation co-design.
Teachers Do More Than Teach: Compressing Image-to-Image Models
In this work, we aim to address these issues by introducing a teacher network that provides a search space in which efficient network architectures can be found, in addition to performing knowledge distillation.
Sanity Checks for Lottery Tickets: Does Your Winning Ticket Really Win the Jackpot?
Based on our analysis, we summarize a guideline for parameter settings in regards of specific architecture characteristics, which we hope to catalyze the research progress on the topic of lottery ticket hypothesis.
SPViT: Enabling Faster Vision Transformers via Soft Token Pruning
Moreover, our framework can guarantee the identified model to meet resource specifications of mobile devices and FPGA, and even achieve the real-time execution of DeiT-T on mobile platforms.
Ranked #4 on
Efficient ViTs
on ImageNet-1K (with DeiT-S)
SparCL: Sparse Continual Learning on the Edge
SparCL achieves both training acceleration and accuracy preservation through the synergy of three aspects: weight sparsity, data efficiency, and gradient sparsity.
MEST: Accurate and Fast Memory-Economic Sparse Training Framework on the Edge
Systematical evaluation on accuracy, training speed, and memory footprint are conducted, where the proposed MEST framework consistently outperforms representative SOTA works.
Peeling the Onion: Hierarchical Reduction of Data Redundancy for Efficient Vision Transformer Training
Vision transformers (ViTs) have recently obtained success in many applications, but their intensive computation and heavy memory usage at both training and inference time limit their generalization.
Pruning-as-Search: Efficient Neural Architecture Search via Channel Pruning and Structural Reparameterization
By combining the structural reparameterization and PaS, we successfully searched out a new family of VGG-like and lightweight networks, which enable the flexibility of arbitrary width with respect to each layer instead of each stage.
Layer Freezing & Data Sieving: Missing Pieces of a Generic Framework for Sparse Training
Therefore, we analyze the feasibility and potentiality of using the layer freezing technique in sparse training and find it has the potential to save considerable training costs.
Self-Ensemble Protection: Training Checkpoints Are Good Data Protectors
In this paper, we uncover them by model checkpoints' gradients, forming the proposed self-ensemble protection (SEP), which is very effective because (1) learning on examples ignored during normal training tends to yield DNNs ignoring normal examples; (2) checkpoints' cross-model gradients are close to orthogonal, meaning that they are as diverse as DNNs with different architectures.
Data Level Lottery Ticket Hypothesis for Vision Transformers
That is, there exists a subset of input image patches such that a ViT can be trained from scratch by using only this subset of patches and achieve similar accuracy to the ViTs trained by using all image patches.
MTS-LOF: Medical Time-Series Representation Learning via Occlusion-Invariant Features
MTS-LOF leverages the strengths of contrastive learning and Masked Autoencoder (MAE) methods, offering a unique approach to representation learning for medical time series data.
Structured Weight Matrices-Based Hardware Accelerators in Deep Neural Networks: FPGAs and ASICs
For FPGA implementations on deep convolutional neural networks (DCNNs), we achieve at least 152X and 72X improvement in performance and energy efficiency, respectively using the SWM-based framework, compared with the baseline of IBM TrueNorth processor under same accuracy constraints using the data set of MNIST, SVHN, and CIFAR-10.
Towards Ultra-High Performance and Energy Efficiency of Deep Learning Systems: An Algorithm-Hardware Co-Optimization Framework
Hardware accelerations of deep learning systems have been extensively investigated in industry and academia.
An Area and Energy Efficient Design of Domain-Wall Memory-Based Deep Convolutional Neural Networks using Stochastic Computing
However, in these works, the memory design optimization is neglected for weight storage, which will inevitably result in large hardware cost.
CirCNN: Accelerating and Compressing Deep Neural Networks Using Block-CirculantWeight Matrices
As the size of DNNs continues to grow, it is critical to improve the energy efficiency and performance while maintaining accuracy.
ResNet Can Be Pruned 60x: Introducing Network Purification and Unused Path Removal (P-RM) after Weight Pruning
The state-of-art DNN structures involve high computation and great demand for memory storage which pose intensive challenge on DNN framework resources.
Toward Extremely Low Bit and Lossless Accuracy in DNNs with Progressive ADMM
Weight quantization is one of the most important techniques of Deep Neural Networks (DNNs) model compression method.
Non-Structured DNN Weight Pruning -- Is It Beneficial in Any Platform?
Based on the proposed comparison framework, with the same accuracy and quantization, the results show that non-structrued pruning is not competitive in terms of both storage and computation efficiency.
Tiny but Accurate: A Pruned, Quantized and Optimized Memristor Crossbar Framework for Ultra Efficient DNN Implementation
To mitigate the challenges, the memristor crossbar array has emerged as an intrinsically suitable matrix computation and low-power acceleration framework for DNN applications.
An Ultra-Efficient Memristor-Based DNN Framework with Structured Weight Pruning and Quantization Using ADMM
Memristor-based weight pruning and weight quantization have been seperately investigated and proven effectiveness in reducing area and power consumption compared to the original DNN model.
A SOT-MRAM-based Processing-In-Memory Engine for Highly Compressed DNN Implementation
Thus, the footprint and power consumption of SOT-MRAM PIM can be reduced, while increasing the overall system throughput at the meantime, making our proposed ADMM-based SOT-MRAM PIM more energy efficiency and suitable for embedded systems or IoT devices.
SS-Auto: A Single-Shot, Automatic Structured Weight Pruning Framework of DNNs with Ultra-High Efficiency
Structured weight pruning is a representative model compression technique of DNNs for hardware efficiency and inference accelerations.
Real-Time Execution of Large-scale Language Models on Mobile
Our framework can guarantee the identified model to meet both resource and real-time specifications of mobile devices, thus achieving real-time execution of large transformer-based models like BERT variants.
ClickTrain: Efficient and Accurate End-to-End Deep Learning Training via Fine-Grained Architecture-Preserving Pruning
Moreover, compared with the state-of-the-art pruning-during-training approach, ClickTrain provides significant improvements both accuracy and compression ratio on the tested CNN models and datasets, under similar limited training time.
NPAS: A Compiler-aware Framework of Unified Network Pruning and Architecture Search for Beyond Real-Time Mobile Acceleration
With the increasing demand to efficiently deploy DNNs on mobile edge devices, it becomes much more important to reduce unnecessary computation and increase the execution speed.
Achieving Real-Time LiDAR 3D Object Detection on a Mobile Device
3D object detection is an important task, especially in the autonomous driving application domain.
Lottery Ticket Preserves Weight Correlation: Is It Desirable or Not?
In deep model compression, the recent finding "Lottery Ticket Hypothesis" (LTH) (Frankle & Carbin, 2018) pointed out that there could exist a winning ticket (i. e., a properly pruned sub-network together with original weight initialization) that can achieve competitive performance than the original dense network.
A Compression-Compilation Framework for On-mobile Real-time BERT Applications
In this paper, we propose a compression-compilation co-design framework that can guarantee the identified model to meet both resource and real-time specifications of mobile devices.
Improving DNN Fault Tolerance using Weight Pruning and Differential Crossbar Mapping for ReRAM-based Edge AI
More importantly, our method does not require extra hardware cost compared to the traditional two-column mapping scheme.
FORMS: Fine-grained Polarized ReRAM-based In-situ Computation for Mixed-signal DNN Accelerator
With weights stored in the ReRAM crossbar cells as conductance, when the input vector is applied to word lines, the matrix-vector multiplication results can be generated as the current in bit lines.
Achieving Real-Time Object Detection on MobileDevices with Neural Pruning Search
Object detection plays an important role in self-driving cars for security development.
Towards Fast and Accurate Multi-Person Pose Estimation on Mobile Devices
The rapid development of autonomous driving, abnormal behavior detection, and behavior recognition makes an increasing demand for multi-person pose estimation-based applications, especially on mobile platforms.
Achieving on-Mobile Real-Time Super-Resolution with Neural Architecture and Pruning Search
Though recent years have witnessed remarkable progress in single image super-resolution (SISR) tasks with the prosperous development of deep neural networks (DNNs), the deep learning methods are confronted with the computation and memory consumption issues in practice, especially for resource-limited platforms such as mobile devices.
Automatic Mapping of the Best-Suited DNN Pruning Schemes for Real-Time Mobile Acceleration
Weight pruning is an effective model compression technique to tackle the challenges of achieving real-time deep neural network (DNN) inference on mobile devices.
Real-Time Portrait Stylization on the Edge
In this work we demonstrate real-time portrait stylization, specifically, translating self-portrait into cartoon or anime style on mobile devices.
Auto-ViT-Acc: An FPGA-Aware Automatic Acceleration Framework for Vision Transformer with Mixed-Scheme Quantization
Compared with state-of-the-art ViT quantization work (algorithmic approach only without hardware acceleration), our quantization achieves 0. 47% to 1. 36% higher Top-1 accuracy under the same bit-width.
Towards Artificial General Intelligence (AGI) in the Internet of Things (IoT): Opportunities and Challenges
Artificial General Intelligence (AGI), possessing the capacity to comprehend, learn, and execute tasks with human cognitive abilities, engenders significant anticipation and intrigue across scientific, commercial, and societal arenas.
SupeRBNN: Randomized Binary Neural Network Using Adiabatic Superconductor Josephson Devices
Specifically, we investigate the randomized behavior of the AQFP devices and analyze the impact of crossbar size on current attenuation, subsequently formulating the current amplitude into the values suitable for use in BNN computation.
Zero-Space Cost Fault Tolerance for Transformer-based Language Models on ReRAM
Resistive Random Access Memory (ReRAM) has emerged as a promising platform for deep neural networks (DNNs) due to its support for parallel in-situ matrix-vector multiplication.
SmartFRZ: An Efficient Training Framework using Attention-Based Layer Freezing
Therefore, there lacks a generic and smart layer freezing method that can automatically perform ``in-situation'' layer freezing for different networks during training processes.
EdgeOL: Efficient in-situ Online Learning on Edge Devices
Emerging applications, such as robot-assisted eldercare and object recognition, generally employ deep learning neural networks (DNNs) and naturally require: i) handling streaming-in inference requests and ii) adapting to possible deployment scenario changes.
Efficient Pruning of Large Language Model with Adaptive Estimation Fusion
Large language models (LLMs) have become crucial for many generative downstream tasks, leading to an inevitable trend and significant challenge to deploy them efficiently on resource-constrained devices.
