SOTAVerified

Knowledge Distillation

Knowledge distillation is the process of transferring knowledge from a large model to a smaller one. While large models (such as very deep neural networks or ensembles of many models) have higher knowledge capacity than small models, this capacity might not be fully utilized.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–100 of 4240 papers

TitleStatusHype
Analyzing the Importance of Blank for CTC-Based Knowledge Distillation—0
Feature Fusion and Knowledge-Distilled Multi-Modal Multi-Target Detection—0
Fine-tune Before Structured Pruning: Towards Compact and Accurate Self-Supervised Models for Speaker Diarization—0
Revisiting Cross-Modal Knowledge Distillation: A Disentanglement Approach for RGBD Semantic SegmentationCode0
Progressive Class-level Distillation—0
CL-LoRA: Continual Low-Rank Adaptation for Rehearsal-Free Class-Incremental LearningCode1
Proactive Guidance of Multi-Turn Conversation in Industrial Search—0
CREFT: Sequential Multi-Agent LLM for Character Relation Extraction—0
A Simple Linear Patch Revives Layer-Pruned Large Language Models—0
Sketch Down the FLOPs: Towards Efficient Networks for Human Sketch—0
Knowledge Distillation for Reservoir-based Classifier: Human Activity Recognition—0
CAST: Contrastive Adaptation and Distillation for Semi-Supervised Instance Segmentation—0
Improving Respiratory Sound Classification with Architecture-Agnostic Knowledge Distillation from EnsemblesCode0
Multi-MLLM Knowledge Distillation for Out-of-Context News Detection—0
EasyDistill: A Comprehensive Toolkit for Effective Knowledge Distillation of Large Language Models—0
Light distillation for Incremental Graph Convolution Collaborative Filtering—0
Model Stitching by Functional Latent Alignment—0
Efficient Speech Translation through Model Compression and Knowledge DistillationCode0
From Data to Modeling: Fully Open-vocabulary Scene Graph Generation—0
DOGe: Defensive Output Generation for LLM Protection Against Knowledge DistillationCode0
ESLM: Risk-Averse Selective Language Modeling for Efficient Pretraining—0
Optimizing edge AI models on HPC systems with the edge in the loopCode0
Mosaic: Data-Free Knowledge Distillation via Mixture-of-Experts for Heterogeneous Distributed EnvironmentsCode0
Online Knowledge Distillation with Reward Guidance—0
Remote Sensing Image Classification with Decoupled Knowledge Distillation—0
Holistic White-light Polyp Classification via Alignment-free Dense Distillation of Auxiliary Optical ChromoendoscopyCode0
Tokenizing Electron Cloud in Protein-Ligand Interaction Learning—0
Knowledge Grafting of Large Language ModelsCode0
C3R: Channel Conditioned Cell Representations for unified evaluation in microscopy imaging—0
Single Snapshot Distillation for Phase Coded Mask Design in Phase Retrieval—0
ToDi: Token-wise Distillation via Fine-Grained Divergence Control—0
On Multilingual Encoder Language Model Compression for Low-Resource Languages—0
SEDD-PCC: A Single Encoder-Dual Decoder Framework For End-To-End Learned Point Cloud Compression—0
MentalMAC: Enhancing Large Language Models for Detecting Mental Manipulation via Multi-Task Anti-Curriculum Distillation—0
Deliberation on Priors: Trustworthy Reasoning of Large Language Models on Knowledge GraphsCode1
On the Generalization vs Fidelity Paradox in Knowledge DistillationCode0
An Efficient Private GPT Never Autoregressively Decodes—0
DeepKD: A Deeply Decoupled and Denoised Knowledge Distillation TrainerCode1
UWSAM: Segment Anything Model Guided Underwater Instance Segmentation and A Large-scale Benchmark DatasetCode1
Bridge the Gap between Past and Future: Siamese Model Optimization for Context-Aware Document Ranking—0
Intra-class Patch Swap for Self-DistillationCode0
Interpretable Traces, Unexpected Outcomes: Investigating the Disconnect in Trace-Based Knowledge Distillation—0
Ground-V: Teaching VLMs to Ground Complex Instructions in Pixels—0
Improved Methods for Model Pruning and Knowledge Distillation—0
Bridging the Modality Gap: Enhancing Channel Prediction with Semantically Aligned LLMs and Knowledge Distillation—0
SMOTExT: SMOTE meets Large Language ModelsCode0
A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank CloneCode1
Towards Low-Latency Event Stream-based Visual Object Tracking: A Slow-Fast ApproachCode0
Robust Multimodal Segmentation with Representation Regularization and Hybrid Prototype DistillationCode0
ORQA: A Benchmark and Foundation Model for Holistic Operating Room Modeling—0
Show:102550
← PrevPage 2 of 85Next →
All datasetsImageNetCIFAR-100COCO (Common Objects in Context)COCO 2017 valPASCAL VOCKITTI

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1ScaleKD (T:BEiT-L S:ViT-B/14)Top-1 accuracy %86.43—Unverified
2ScaleKD (T:Swin-L S:ViT-B/16)Top-1 accuracy %85.53—Unverified
3ScaleKD (T:Swin-L S:ViT-S/16)Top-1 accuracy %83.93—Unverified
4ScaleKD (T:Swin-L S:Swin-T)Top-1 accuracy %83.8—Unverified
5KD++(T: regnety-16GF S:ViT-B)Top-1 accuracy %83.6—Unverified
6VkD (T:RegNety 160 S:DeiT-S)Top-1 accuracy %82.9—Unverified
7SpectralKD (T:Swin-S S:Swin-T)Top-1 accuracy %82.7—Unverified
8ScaleKD (T:Swin-L S:ResNet-50)Top-1 accuracy %82.55—Unverified
9DiffKD (T:Swin-L S: Swin-T)Top-1 accuracy %82.5—Unverified
10DIST (T: Swin-L S: Swin-T)Top-1 accuracy %82.3—Unverified
#ModelMetricClaimedVerifiedStatus
1SRD (T:resnet-32x4, S:shufflenet-v2)Top-1 Accuracy (%)79.86—Unverified
2shufflenet-v2(T:resnet-32x4, S:shufflenet-v2)Top-1 Accuracy (%)78.76—Unverified
3MV-MR (T: CLIP/ViT-B-16 S: resnet50)Top-1 Accuracy (%)78.6—Unverified
4resnet8x4 (T: resnet32x4 S: resnet8x4)Top-1 Accuracy (%)78.28—Unverified
5resnet8x4 (T: resnet32x4 S: resnet8x4 [modified])Top-1 Accuracy (%)78.08—Unverified
6ReviewKD++(T:resnet-32x4, S:shufflenet-v2)Top-1 Accuracy (%)77.93—Unverified
7ReviewKD++(T:resnet-32x4, S:shufflenet-v1)Top-1 Accuracy (%)77.68—Unverified
8resnet8x4 (T: resnet32x4 S: resnet8x4)Top-1 Accuracy (%)77.5—Unverified
9resnet8x4 (T: resnet32x4 S: resnet8x4)Top-1 Accuracy (%)76.68—Unverified
10resnet8x4 (T: resnet32x4 S: resnet8x4)Top-1 Accuracy (%)76.31—Unverified
#ModelMetricClaimedVerifiedStatus
1LSHFM (T: ResNet101 S: ResNet50)mAP77.16—Unverified
2LSHFM (T: ResNet101 S: MobileNetV2)mAP73.73—Unverified
3ADLIK-Faster (T: Faster R-CNN vit-base S: Faster R-CNN deit-small) box AP47.6—Unverified
4ADLIK-Mask (T: Mask R-CNN vit-base S: Mask R-CNN deit-small)mask AP42.4—Unverified
#ModelMetricClaimedVerifiedStatus
1ReviewKD++(T: faster rcnn(resnet101), S:faster rcnn(resnet50))[email protected]61.8—Unverified
2ReviewKD++(T: faster rcnn(resnet101), S:faster rcnn(resnet18))[email protected]57.96—Unverified
3ReviewKD++(T: faster rcnn(resnet101), S:faster rcnn(mobilenet-v2))[email protected]55.18—Unverified
#ModelMetricClaimedVerifiedStatus
1LSHFM (T: ResNet101 S: ResNet50)mAP93.17—Unverified
2LSHFM (T: ResNet101 S: MobileNetV2)mAP90.14—Unverified
#ModelMetricClaimedVerifiedStatus
1TIE-KD (T: Adabins S: MobileNetV2)RMSE2.43—Unverified