Accuracy Improvement

Accuracy improvement in machine learning is a central research theme focused on enhancing the reliability and performance of various models across diverse applications. Current efforts concentrate on refining existing architectures like transformers and convolutional neural networks, developing novel techniques such as feedback mechanisms and dynamic model switching, and leveraging strategies like knowledge transfer and synthetic data augmentation. These advancements are crucial for improving the dependability of AI systems in critical domains like healthcare, finance, and autonomous systems, ultimately leading to more robust and efficient solutions.

Papers

February 21, 2023

AttentionMixer: An Accurate and Interpretable Framework for Process Monitoring
Hao Wang, Zhiyu Wang, Yunlong Niu, Zhaoran Liu, Haozhe Li, Yilin Liao, Yuxin Huang, Xinggao Liu
Accuracy Improvement Process Monitoring

February 1, 2023

Improved Exact and Heuristic Algorithms for Maximum Weight Clique
Roman Erhardt, Kathrin Hanauer, Nils Kriege, Christian Schulz, Darren Strash
Graph Structure Real World Graph Accuracy Improvement Local Graph Heuristic Algorithm Maximum Clique Reduced Graph

November 19, 2022

Accuracy Booster: Enabling 4-bit Fixed-point Arithmetic for DNN Training
Simla Burcu Harma, Ayan Chakraborty, Nicholas Sperry, Babak Falsafi, Martin Jaggi, Yunho Oh
Low Precision Accuracy Improvement DNN Training Bit Training

October 28, 2022

ROMA: Run-Time Object Detection To Maximize Real-Time Accuracy
JunKyu Lee, Blesson Varghese, Hans Vandierendonck
Object Detection Accuracy Improvement YOLOv5 Object

October 13, 2022

AccelAT: A Framework for Accelerating the Adversarial Training of Deep Neural Networks through Accuracy Gradient
Farzad Nikfam, Alberto Marchisio, Maurizio Martina, Muhammad Shafique
Deep Neural Network Adversarial Attack New Framework Adversarial Training Robust Deep Accuracy Improvement

September 26, 2022

Efficient Multi-Prize Lottery Tickets: Enhanced Accuracy, Training, and Inference Speed
Hao Cheng, Pu Zhao, Yize Li, Xue Lin, James Diffenderfer, Ryan Goldhahn, Bhavya Kailkhura
Neural Network Training Data High Efficiency Multidimensional Local Precision Rate Binary Neural Network Accuracy Improvement Lottery Ticket Inference Speed Inference Speedup Pruning Ratio

August 17, 2022

Random Search Hyper-Parameter Tuning: Expected Improvement Estimation and the Corresponding Lower Bound
Dan Navon, Alex M. Bronstein
Hyperparameter Optimization Lower Bound Hyperparameter Tuning Related Hyperparameters Local Search Accuracy Improvement Hyper Parameter Tuning Expected Improvement

August 3, 2022

Centroids Matching: an efficient Continual Learning approach operating in the embedding space
Jary Pomponi, Simone Scardapane, Aurelio Uncini
Catastrophic Forgetting Incremental Learning Deep Space Feature Space Accuracy Improvement Efficient Continual Learning Novel Regularization Feature Centroid

July 1, 2022

Visual Transformer Meets CutMix for Improved Accuracy, Communication Efficiency, and Data Privacy in Split Learning
Sihun Baek, Jihong Park, Praneeth Vepakomma, Ramesh Raskar, Mehdi Bennis, Seong-Lyun Kim
Convolutional Neural Network Data Augmentation Split Learning Layer Wise Communication Efficiency Accuracy Improvement Visual Transformer

June 15, 2022

Differentiable Top-k Classification Learning
Felix Petersen, Hilde Kuehne, Christian Borgelt, Oliver Deussen
Accuracy Improvement Top $K$ Classification Differentiable Sorting

April 14, 2022

Machine Learning-Based Automated Thermal Comfort Prediction: Integration of Low-Cost Thermal and Visual Cameras for Higher Accuracy
Roshanak Ashrafi, Mona Azarbayjani, Hamed Tabkhi
New Machine Thermal Image Visual Sensor Accuracy Improvement Thermal Camera

April 2, 2022

Class-Incremental Learning by Knowledge Distillation with Adaptive Feature Consolidation
Minsoo Kang, Jaeyoo Park, Bohyung Han
Deep Neural Network Knowledge Distillation Catastrophic Forgetting Incremental Learning Class Incremental Learning Accuracy Improvement

March 11, 2022

QDrop: Randomly Dropping Quantization for Extremely Low-bit Post-Training Quantization
Xiuying Wei, Ruihao Gong, Yuhang Li, Xianglong Liu, Fengwei Yu
Quantization Operator Multiplier Free Quantization Accuracy Improvement Activation Quantization Low Bit

March 4, 2022

Exploration of Various Deep Learning Models for Increased Accuracy in Automatic Polyp Detection
Ariel E. Isidro, Arnel C. Fajardo, Alexander A. Hernandez
Deep Learning Deep Learning Model Transfer Learning Environment Exploration Accuracy Improvement Polyp Detection Different Convolutional Neural Network Colonoscopy Image

February 24, 2022

Fourier-Based Augmentations for Improved Robustness and Uncertainty Calibration
Ryan Soklaski, Michael Yee, Theodoros Tsiligkaridis
Data Augmentation Model Robustness Accuracy Improvement Uncertainty Calibration Better Robustness

November 12, 2021

BSC: Block-based Stochastic Computing to Enable Accurate and Efficient TinyML
Yuhong Song, Edwin Hsing-Mean Sha, Qingfeng Zhuge, Rui Xu, Yongzhuo Zhang, Bingzhe Li, Lei Yang
Machine Learning TinyML Model Accuracy Improvement Low Power Stochastic Computing

Accuracy Improvement

Papers

AttentionMixer: An Accurate and Interpretable Framework for Process Monitoring

Improved Exact and Heuristic Algorithms for Maximum Weight Clique

Accuracy Booster: Enabling 4-bit Fixed-point Arithmetic for DNN Training

ROMA: Run-Time Object Detection To Maximize Real-Time Accuracy

AccelAT: A Framework for Accelerating the Adversarial Training of Deep Neural Networks through Accuracy Gradient

Efficient Multi-Prize Lottery Tickets: Enhanced Accuracy, Training, and Inference Speed

Random Search Hyper-Parameter Tuning: Expected Improvement Estimation and the Corresponding Lower Bound

Centroids Matching: an efficient Continual Learning approach operating in the embedding space

Visual Transformer Meets CutMix for Improved Accuracy, Communication Efficiency, and Data Privacy in Split Learning

Differentiable Top-k Classification Learning

Machine Learning-Based Automated Thermal Comfort Prediction: Integration of Low-Cost Thermal and Visual Cameras for Higher Accuracy

Class-Incremental Learning by Knowledge Distillation with Adaptive Feature Consolidation

QDrop: Randomly Dropping Quantization for Extremely Low-bit Post-Training Quantization

Exploration of Various Deep Learning Models for Increased Accuracy in Automatic Polyp Detection

Fourier-Based Augmentations for Improved Robustness and Uncertainty Calibration

BSC: Block-based Stochastic Computing to Enable Accurate and Efficient TinyML