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 151–200 of 1356 papers

TitleStatusHype
Enhancing Cross-Tokenizer Knowledge Distillation with Contextual Dynamical MappingCode1
Environmental Sound Classification on the Edge: A Pipeline for Deep Acoustic Networks on Extremely Resource-Constrained DevicesCode1
Examining Post-Training Quantization for Mixture-of-Experts: A BenchmarkCode1
Contrastive Representation DistillationCode1
Aligned Structured Sparsity Learning for Efficient Image Super-ResolutionCode1
FAT: Learning Low-Bitwidth Parametric Representation via Frequency-Aware TransformationCode1
FFNeRV: Flow-Guided Frame-Wise Neural Representations for VideosCode1
FIMA-Q: Post-Training Quantization for Vision Transformers by Fisher Information Matrix ApproximationCode1
Consistent Quantity-Quality Control across Scenes for Deployment-Aware Gaussian SplattingCode1
Computation-Efficient Knowledge Distillation via Uncertainty-Aware MixupCode1
Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer InferenceCode1
Topology-Aware Network Pruning using Multi-stage Graph Embedding and Reinforcement LearningCode1
Head Network Distillation: Splitting Distilled Deep Neural Networks for Resource-Constrained Edge Computing SystemsCode1
HiNeRV: Video Compression with Hierarchical Encoding-based Neural RepresentationCode1
Hyper-Compression: Model Compression via HyperfunctionCode1
Implicit Regularization via Neural Feature AlignmentCode1
CPrune: Compiler-Informed Model Pruning for Efficient Target-Aware DNN ExecutionCode1
Initialization and Regularization of Factorized Neural LayersCode1
Streamlining Redundant Layers to Compress Large Language ModelsCode1
KD-MRI: A knowledge distillation framework for image reconstruction and image restoration in MRI workflowCode1
Activation-Informed Merging of Large Language ModelsCode1
A Winning Hand: Compressing Deep Networks Can Improve Out-Of-Distribution RobustnessCode1
Accurate Retraining-free Pruning for Pretrained Encoder-based Language ModelsCode1
Backdoor Attacks on Federated Learning with Lottery Ticket HypothesisCode1
Compression-Aware Video Super-ResolutionCode1
Less is More: Task-aware Layer-wise Distillation for Language Model CompressionCode1
Composable Interventions for Language ModelsCode1
Basis Sharing: Cross-Layer Parameter Sharing for Large Language Model CompressionCode1
LiMuSE: Lightweight Multi-modal Speaker ExtractionCode1
LiteYOLO-ID: A Lightweight Object Detection Network for Insulator Defect DetectionCode1
Compacting, Picking and Growing for Unforgetting Continual LearningCode1
Localize-and-Stitch: Efficient Model Merging via Sparse Task ArithmeticCode1
Comprehensive Knowledge Distillation with Causal InterventionCode1
BERT-EMD: Many-to-Many Layer Mapping for BERT Compression with Earth Mover's DistanceCode1
BERT-of-Theseus: Compressing BERT by Progressive Module ReplacingCode1
Memory-Efficient Backpropagation through Large Linear LayersCode1
CompRess: Self-Supervised Learning by Compressing RepresentationsCode1
Mix-LN: Unleashing the Power of Deeper Layers by Combining Pre-LN and Post-LNCode1
MobileNMT: Enabling Translation in 15MB and 30msCode1
Bidirectional Distillation for Top-K Recommender SystemCode1
Clustered Sampling: Low-Variance and Improved Representativity for Clients Selection in Federated LearningCode1
CoA: Towards Real Image Dehazing via Compression-and-AdaptationCode1
Online Knowledge Distillation via Collaborative LearningCode1
Bit-mask Robust Contrastive Knowledge Distillation for Unsupervised Semantic HashingCode1
An Efficient Multilingual Language Model Compression through Vocabulary TrimmingCode1
Model LEGO: Creating Models Like Disassembling and Assembling Building BlocksCode1
Passport-aware Normalization for Deep Model ProtectionCode1
Paying more attention to snapshots of Iterative Pruning: Improving Model Compression via Ensemble DistillationCode1
CrossKD: Cross-Head Knowledge Distillation for Object DetectionCode1
Distilling Linguistic Context for Language Model CompressionCode1
Show:102550
← PrevPage 4 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