SOTAVerified

Model Compression

Model Compression is an actively pursued area of research over the last few years with the goal of deploying state-of-the-art deep networks in low-power and resource limited devices without significant drop in accuracy. Parameter pruning, low-rank factorization and weight quantization are some of the proposed methods to compress the size of deep networks.

Source: KD-MRI: A knowledge distillation framework for image reconstruction and image restoration in MRI workflow

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 851–900 of 1356 papers

TitleStatusHype
Data-Free Knowledge Distillation for Image Super-ResolutionCode0
Quantized Neural Networks via -1, +1 Encoding Decomposition and AccelerationCode0
How does topology of neural architectures impact gradient propagation and model performance?Code0
Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and BetterCode1
Topology Distillation for Recommender System—0
Masked Training of Neural Networks with Partial Gradients—0
A Winning Hand: Compressing Deep Networks Can Improve Out-Of-Distribution RobustnessCode1
Energy-efficient Knowledge Distillation for Spiking Neural Networks—0
Heterogeneous Federated Learning using Dynamic Model Pruning and Adaptive Gradient—0
ModelDiff: Testing-Based DNN Similarity Comparison for Model Reuse DetectionCode1
FedNILM: Applying Federated Learning to NILM Applications at the Edge—0
FedNL: Making Newton-Type Methods Applicable to Federated Learning—0
Feature Flow Regularization: Improving Structured Sparsity in Deep Neural Networks—0
Bidirectional Distillation for Top-K Recommender SystemCode1
You Only Compress Once: Towards Effective and Elastic BERT Compression via Exploit-Explore Stochastic Nature GradientCode1
Enabling Lightweight Fine-tuning for Pre-trained Language Model Compression based on Matrix Product OperatorsCode1
One Teacher is Enough? Pre-trained Language Model Distillation from Multiple Teachers—0
Energy-Efficient Model Compression and Splitting for Collaborative Inference Over Time-Varying Channels—0
On Attention Redundancy: A Comprehensive Study—0
NAS-BERT: Task-Agnostic and Adaptive-Size BERT Compression with Neural Architecture Search—0
Towards Efficient Full 8-bit Integer DNN Online Training on Resource-limited Devices without Batch Normalization—0
Super Tickets in Pre-Trained Language Models: From Model Compression to Improving GeneralizationCode1
Differentiable Sparsification for Deep Neural Networks—0
Model Compression—0
How to Explain Neural Networks: an Approximation Perspective—0
Clustered Sampling: Low-Variance and Improved Representativity for Clients Selection in Federated LearningCode1
3U-EdgeAI: Ultra-Low Memory Training, Ultra-Low BitwidthQuantization, and Ultra-Low Latency Acceleration—0
Test-Time Adaptation Toward Personalized Speech Enhancement: Zero-Shot Learning with Knowledge Distillation—0
Neural 3D Scene Compression via Model Compression—0
Encoding Weights of Irregular Sparsity for Fixed-to-Fixed Model Compression—0
Modulating Regularization Frequency for Efficient Compression-Aware Model Training—0
Initialization and Regularization of Factorized Neural LayersCode1
Knowledge Distillation for Swedish NER models: A Search for Performance and Efficiency—0
On the Adversarial Robustness of Quantized Neural Networks—0
Stealthy Backdoors as Compression ArtifactsCode0
Spirit Distillation: A Model Compression Method with Multi-domain Knowledge Transfer—0
Spatio-Temporal Pruning and Quantization for Low-latency Spiking Neural Networks—0
Skip-Convolutions for Efficient Video ProcessingCode1
Knowledge Distillation as Semiparametric InferenceCode0
Differentiable Model Compression via Pseudo Quantization NoiseCode1
Compact CNN Structure Learning by Knowledge Distillation—0
Augmenting Deep Classifiers with Polynomial Neural NetworksCode0
Annealing Knowledge DistillationCode0
Dual Discriminator Adversarial Distillation for Data-free Model Compression—0
Reversible Watermarking in Deep Convolutional Neural Networks for Integrity Authentication—0
Efficient Personalized Speech Enhancement through Self-Supervised Learning—0
Model Compression for Dynamic Forecast CombinationCode0
Tight Compression: Compressing CNN Through Fine-Grained Pruning and Weight Permutation for Efficient Implementation—0
Deep Compression for PyTorch Model Deployment on MicrocontrollersCode1
Shrinking Bigfoot: Reducing wav2vec 2.0 footprint—0
Show:102550
← PrevPage 18 of 28Next →
All datasetsImageNetQNLI

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1ADLIK-MO-ResNet50+W4A4Top-177.88—Unverified
2ADLIK-MO-ResNet50+W3A4Top-177.34—Unverified
3ResNet-18 + 4bit-1dim model compression using DKMTop-170.52—Unverified
4MobileNet-v1 + 4bit-1dim model compression using DKMTop-169.63—Unverified
5ResNet-18 + 2bit-1dim model compression using DKMTop-168.63—Unverified
6MobileNet-v1 + 2bit-1dim model compression using DKMTop-167.62—Unverified
7ResNet-18 + 4bit-4dim model compression using DKMTop-166.1—Unverified
8ResNet-18 + 2bit-2dim model compression using DKMTop-164.7—Unverified
9MobileNet-v1 + 4bit-4dim model compression using DKMTop-161.4—Unverified
10ResNet-18 + 1bit-1dim model compression using DKMTop-159.7—Unverified
#ModelMetricClaimedVerifiedStatus
1MobileBERT + 2bit-1dim model compression using DKMAccuracy82.13—Unverified
2MobileBERT + 1bit-1dim model compression using DKMAccuracy63.17—Unverified