Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Marios Savvides

Marios Savvides

Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study


Apr 01, 2021
Zhiqiang Shen, Zechun Liu, Dejia Xu, Zitian Chen, Kwang-Ting Cheng, Marios Savvides

Figure 1 for Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study
Figure 2 for Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study
Figure 3 for Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study
Figure 4 for Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study
* ICLR 2021. Project page: http://zhiqiangshen.com/projects/LS_and_KD/index.html 
  Access Paper or Ask Questions

Unsupervised Disentanglement of Linear-Encoded Facial Semantics


Mar 30, 2021
Yutong Zheng, Yu-Kai Huang, Ran Tao, Zhiqiang Shen, Marios Savvides

Figure 1 for Unsupervised Disentanglement of Linear-Encoded Facial Semantics
Figure 2 for Unsupervised Disentanglement of Linear-Encoded Facial Semantics
Figure 3 for Unsupervised Disentanglement of Linear-Encoded Facial Semantics
Figure 4 for Unsupervised Disentanglement of Linear-Encoded Facial Semantics
* Accepted in IEEE Conference on Computer Vision and Pattern Recognition 2021 (CVPR2021) 
  Access Paper or Ask Questions

Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection


Mar 19, 2021
Chenchen Zhu, Fangyi Chen, Uzair Ahmed, Zhiqiang Shen, Marios Savvides

Figure 1 for Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection
Figure 2 for Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection
Figure 3 for Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection
Figure 4 for Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection
* CVPR 2021 
  Access Paper or Ask Questions

S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration


Feb 17, 2021
Zhiqiang Shen, Zechun Liu, Jie Qin, Lei Huang, Kwang-Ting Cheng, Marios Savvides

Figure 1 for S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration
Figure 2 for S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration
Figure 3 for S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration
Figure 4 for S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration
* Technical report, 10 pages. Self-supervised binary neural networks using distillation loss (5.5~15% improvement over contrastive baseline) 
  Access Paper or Ask Questions

Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning


Feb 08, 2021
Zhiqiang Shen, Zechun Liu, Jie Qin, Marios Savvides, Kwang-Ting Cheng

Figure 1 for Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning
Figure 2 for Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning
Figure 3 for Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning
Figure 4 for Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning
* AAAI 2021. A search based fine-tuning strategy for few-shot learning 
  Access Paper or Ask Questions

A Multi-task Contextual Atrous Residual Network for Brain Tumor Detection & Segmentation


Dec 03, 2020
Ngan Le, Kashu Yamazaki, Dat Truong, Kha Gia Quach, Marios Savvides

Figure 1 for A Multi-task Contextual Atrous Residual Network for Brain Tumor Detection & Segmentation
Figure 2 for A Multi-task Contextual Atrous Residual Network for Brain Tumor Detection & Segmentation
Figure 3 for A Multi-task Contextual Atrous Residual Network for Brain Tumor Detection & Segmentation
Figure 4 for A Multi-task Contextual Atrous Residual Network for Brain Tumor Detection & Segmentation
* Accepted in ICPR 2020 
  Access Paper or Ask Questions

Online Ensemble Model Compression using Knowledge Distillation


Nov 15, 2020
Devesh Walawalkar, Zhiqiang Shen, Marios Savvides

Figure 1 for Online Ensemble Model Compression using Knowledge Distillation
Figure 2 for Online Ensemble Model Compression using Knowledge Distillation
Figure 3 for Online Ensemble Model Compression using Knowledge Distillation
Figure 4 for Online Ensemble Model Compression using Knowledge Distillation

  Access Paper or Ask Questions

MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks


Sep 17, 2020
Zhiqiang Shen, Marios Savvides

Figure 1 for MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks
Figure 2 for MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks
Figure 3 for MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks
* A short technical report. Code and pre-trained models are available at: https://github.com/szq0214/MEAL-V2 
  Access Paper or Ask Questions

Binarizing MobileNet via Evolution-based Searching


May 15, 2020
Hai Phan, Zechun Liu, Dang Huynh, Marios Savvides, Kwang-Ting Cheng, Zhiqiang Shen

Figure 1 for Binarizing MobileNet via Evolution-based Searching
Figure 2 for Binarizing MobileNet via Evolution-based Searching
Figure 3 for Binarizing MobileNet via Evolution-based Searching
Figure 4 for Binarizing MobileNet via Evolution-based Searching
* Accepted by CVPR2020 
  Access Paper or Ask Questions

Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification


Apr 05, 2020
Devesh Walawalkar, Zhiqiang Shen, Zechun Liu, Marios Savvides

Figure 1 for Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification
Figure 2 for Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification
Figure 3 for Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification
Figure 4 for Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification

  Access Paper or Ask Questions

Rethinking Image Mixture for Unsupervised Visual Representation Learning


Mar 11, 2020
Zhiqiang Shen, Zechun Liu, Zhuang Liu, Marios Savvides, Trevor Darrell

Figure 1 for Rethinking Image Mixture for Unsupervised Visual Representation Learning
Figure 2 for Rethinking Image Mixture for Unsupervised Visual Representation Learning
Figure 3 for Rethinking Image Mixture for Unsupervised Visual Representation Learning
Figure 4 for Rethinking Image Mixture for Unsupervised Visual Representation Learning

  Access Paper or Ask Questions

ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions


Mar 07, 2020
Zechun Liu, Zhiqiang Shen, Marios Savvides, Kwang-Ting Cheng

Figure 1 for ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions
Figure 2 for ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions
Figure 3 for ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions
Figure 4 for ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions

  Access Paper or Ask Questions

Solving Missing-Annotation Object Detection with Background Recalibration Loss


Feb 12, 2020
Han Zhang, Fangyi Chen, Zhiqiang Shen, Qiqi Hao, Chenchen Zhu, Marios Savvides

Figure 1 for Solving Missing-Annotation Object Detection with Background Recalibration Loss
Figure 2 for Solving Missing-Annotation Object Detection with Background Recalibration Loss
Figure 3 for Solving Missing-Annotation Object Detection with Background Recalibration Loss
Figure 4 for Solving Missing-Annotation Object Detection with Background Recalibration Loss
* 5 pages. Paper has been accepted by ICASSP 2020 for presentation in a lecture (oral) session 
  Access Paper or Ask Questions

Soft Anchor-Point Object Detection


Nov 27, 2019
Chenchen Zhu, Fangyi Chen, Zhiqiang Shen, Marios Savvides

Figure 1 for Soft Anchor-Point Object Detection
Figure 2 for Soft Anchor-Point Object Detection
Figure 3 for Soft Anchor-Point Object Detection
Figure 4 for Soft Anchor-Point Object Detection

  Access Paper or Ask Questions

Towards a Hypothesis on Visual Transformation based Self-Supervision


Nov 24, 2019
Dipan K. Pal, Sreena Nallamothu, Marios Savvides

Figure 1 for Towards a Hypothesis on Visual Transformation based Self-Supervision
Figure 2 for Towards a Hypothesis on Visual Transformation based Self-Supervision
Figure 3 for Towards a Hypothesis on Visual Transformation based Self-Supervision
Figure 4 for Towards a Hypothesis on Visual Transformation based Self-Supervision
* Draft 
  Access Paper or Ask Questions

SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses


Nov 21, 2019
Zhiqiang Shen, Harsh Maheshwari, Weichen Yao, Marios Savvides

Figure 1 for SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses
Figure 2 for SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses
Figure 3 for SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses
Figure 4 for SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses

  Access Paper or Ask Questions

Learning Non-Parametric Invariances from Data with Permanent Random Connectomes


Nov 13, 2019
Dipan K. Pal, Marios Savvides

Figure 1 for Learning Non-Parametric Invariances from Data with Permanent Random Connectomes
Figure 2 for Learning Non-Parametric Invariances from Data with Permanent Random Connectomes
* Preprint (accepted at NeurIPS SVRHM 2019) 
  Access Paper or Ask Questions

Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement


Aug 22, 2019
Zhiqiang Shen, Zhankui He, Wanyun Cui, Jiahui Yu, Yutong Zheng, Chenchen Zhu, Marios Savvides

Figure 1 for Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement
Figure 2 for Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement
Figure 3 for Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement
Figure 4 for Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement
* This is an extended version of our previous conference paper arXiv:1812.02425 
  Access Paper or Ask Questions

MoBiNet: A Mobile Binary Network for Image Classification


Jul 31, 2019
Hai Phan, Dang Huynh, Yihui He, Marios Savvides, Zhiqiang Shen

Figure 1 for MoBiNet: A Mobile Binary Network for Image Classification
Figure 2 for MoBiNet: A Mobile Binary Network for Image Classification
Figure 3 for MoBiNet: A Mobile Binary Network for Image Classification
Figure 4 for MoBiNet: A Mobile Binary Network for Image Classification

  Access Paper or Ask Questions

Proximal Splitting Networks for Image Restoration


Mar 17, 2019
Raied Aljadaany, Dipan K. Pal, Marios Savvides

Figure 1 for Proximal Splitting Networks for Image Restoration
Figure 2 for Proximal Splitting Networks for Image Restoration
Figure 3 for Proximal Splitting Networks for Image Restoration
Figure 4 for Proximal Splitting Networks for Image Restoration

  Access Paper or Ask Questions

Feature Selective Anchor-Free Module for Single-Shot Object Detection


Mar 02, 2019
Chenchen Zhu, Yihui He, Marios Savvides

Figure 1 for Feature Selective Anchor-Free Module for Single-Shot Object Detection
Figure 2 for Feature Selective Anchor-Free Module for Single-Shot Object Detection
Figure 3 for Feature Selective Anchor-Free Module for Single-Shot Object Detection
Figure 4 for Feature Selective Anchor-Free Module for Single-Shot Object Detection
* CVPR 2019 
  Access Paper or Ask Questions

RankGAN: A Maximum Margin Ranking GAN for Generating Faces


Dec 19, 2018
Rahul Dey, Felix Juefei-Xu, Vishnu Naresh Boddeti, Marios Savvides

Figure 1 for RankGAN: A Maximum Margin Ranking GAN for Generating Faces
Figure 2 for RankGAN: A Maximum Margin Ranking GAN for Generating Faces
Figure 3 for RankGAN: A Maximum Margin Ranking GAN for Generating Faces
Figure 4 for RankGAN: A Maximum Margin Ranking GAN for Generating Faces
* Best Student Paper Award at Asian Conference on Computer Vision (ACCV), 2018 at Perth, Australia. Includes main paper and supplementary material. Total 32 pages including references 
  Access Paper or Ask Questions

Deep Recurrent Level Set for Segmenting Brain Tumors


Oct 10, 2018
T. Hoang Ngan Le, Raajitha Gummadi, Marios Savvides

Figure 1 for Deep Recurrent Level Set for Segmenting Brain Tumors
Figure 2 for Deep Recurrent Level Set for Segmenting Brain Tumors
Figure 3 for Deep Recurrent Level Set for Segmenting Brain Tumors
Figure 4 for Deep Recurrent Level Set for Segmenting Brain Tumors
* booktitle="Medical Image Computing and Computer Assisted Intervention -- MICCAI 2018", year="2018", publisher="Springer International Publishing", 
  Access Paper or Ask Questions

Softer-NMS: Rethinking Bounding Box Regression for Accurate Object Detection


Sep 23, 2018
Yihui He, Xiangyu Zhang, Marios Savvides, Kris Kitani

Figure 1 for Softer-NMS: Rethinking Bounding Box Regression for Accurate Object Detection
Figure 2 for Softer-NMS: Rethinking Bounding Box Regression for Accurate Object Detection
Figure 3 for Softer-NMS: Rethinking Bounding Box Regression for Accurate Object Detection
Figure 4 for Softer-NMS: Rethinking Bounding Box Regression for Accurate Object Detection

  Access Paper or Ask Questions

Non-Parametric Transformation Networks


Sep 08, 2018
Dipan K. Pal, Marios Savvides

Figure 1 for Non-Parametric Transformation Networks
Figure 2 for Non-Parametric Transformation Networks
Figure 3 for Non-Parametric Transformation Networks
Figure 4 for Non-Parametric Transformation Networks
* Preprint only 
  Access Paper or Ask Questions