Self Training Pipeline

Self-training pipelines leverage unlabeled data to enhance the performance of machine learning models, primarily addressing the limitations of supervised learning where labeled data is scarce, expensive, or biased. Current research focuses on improving the accuracy and robustness of self-training through techniques like bilevel optimization, contrastive learning, and consistency regularization, often implemented within architectures such as ResNet, and applied to diverse tasks including image classification, segmentation, and optical flow estimation. These advancements are significant because they enable the development of more accurate and generalizable models, particularly in domains with limited labeled data, such as medical image analysis and remote sensing.

Papers

October 3, 2024

BiSSL: Bilevel Optimization for Self-Supervised Pre-Training and Fine-Tuning
Gustav Wagner Zakarias, Lars Kai Hansen, Zheng-Hua Tan
Self Supervised Learning Fine Tuning Bilevel Optimization Self Supervised Pre Training Training Algorithm Image Classification Datasets Self Training Pipeline

July 14, 2024

A Self-Supervised Learning Pipeline for Demographically Fair Facial Attribute Classification
Sreeraj Ramachandran, Ajita Rattani
Self Supervised Learning Face Recognition Supervised Learning Bias Mitigation Facial Attribute Self Training Pipeline

January 14, 2024

Semi-Supervised Semantic Segmentation using Redesigned Self-Training for White Blood Cells
Vinh Quoc Luu, Duy Khanh Le, Huy Thanh Nguyen, Minh Thanh Nguyen, Thinh Tien Nguyen, Vinh Quang Dinh
Self Training Semi Supervised Semantic Segmentation Self Training Pipeline White Blood Cell

October 3, 2023

MUSCLE: Multi-task Self-supervised Continual Learning to Pre-train Deep Models for X-ray Images of Multiple Body Parts
Weibin Liao, Haoyi Xiong, Qingzhong Wang, Yan Mo, Xuhong Li, Yi Liu, Zeyu Chen, Siyu Huang, Dejing Dou
Pre Trained Model Deep Model X Ray X Ray Image Continual Pre Training Body Part Medical Imaging Task Self Training Pipeline Self Supervised Continual Learning

March 10, 2023

Self-NeRF: A Self-Training Pipeline for Few-Shot Neural Radiance Fields
Jiayang Bai, Letian Huang, Wen Gong, Jie Guo, Yanwen Guo
Neural Radiance Field Radiance Field Self Training Pipeline

March 9, 2023

Taming Contrast Maximization for Learning Sequential, Low-latency, Event-based Optical Flow
Federico Paredes-Vallés, Kirk Y. W. Scheper, Christophe De Wagter, Guido C. H. E. de Croon
Optical Flow Event Camera Low Latency Sequential Learning Event Based Optical Flow Self Training Pipeline Contrast Maximization

February 7, 2023

Explainable Action Prediction through Self-Supervision on Scene Graphs
Pawit Kochakarn, Daniele De Martini, Daniel Omeiza, Lars Kunze
Autonomous Driving Scene Graph Self Supervision Attention Based Architecture Self Training Pipeline Driver Intention

January 18, 2023

Enhancing Self-Training Methods
Aswathnarayan Radhakrishnan, Jim Davis, Zachary Rabin, Benjamin Lewis, Matthew Scherreik, Roman Ilin
Semi Supervised Pseudo Label Self Training Self Training Pipeline Semi Supervised Teacher Student

December 28, 2022

Demonstrate-Search-Predict: Composing retrieval and language models for knowledge-intensive NLP
Omar Khattab, Keshav Santhanam, Xiang Lisa Li, David Hall, Percy Liang, Christopher Potts, Matei Zaharia
Language Model App to App Retrieval Retrieval Model Knowledge Intensive Frozen Language Model Self Training Pipeline Retrieval Augmented in Context Learning Visual Predictive

November 4, 2022

1Cademy @ Causal News Corpus 2022: Leveraging Self-Training in Causality Classification of Socio-Political Event Data
Adam Nik, Ge Zhang, Xingran Chen, Mingyu Li, Jie Fu
Self Training Event Causality Identification Causal Categorization Self Training Pipeline Negative Pseudo Label Causal News Corpus

July 21, 2022

KD-MVS: Knowledge Distillation Based Self-supervised Learning for Multi-view Stereo
Yikang Ding, Qingtian Zhu, Xiangyue Liu, Wentao Yuan, Haotian Zhang, Chi Zhang
Knowledge Distillation Self Supervised Multi View Multi View Stereo Ground Truth Depth Self Training Pipeline

January 20, 2022

CELESTIAL: Classification Enabled via Labelless Embeddings with Self-supervised Telescope Image Analysis Learning
Suhas Kotha, Anirudh Koul, Siddha Ganju, Meher Kasam
Classification Code Label Free Scene Classification Self Training Pipeline

December 2, 2021

Multi-task Self-distillation for Graph-based Semi-Supervised Learning
Yating Ren, Junzhong Ji, Lingfeng Niu, Minglong Lei
Convolutional Neural Network Self Supervised Learning Semi Supervised Self Distillation Self Training Pipeline

Self Training Pipeline

Papers

BiSSL: Bilevel Optimization for Self-Supervised Pre-Training and Fine-Tuning

A Self-Supervised Learning Pipeline for Demographically Fair Facial Attribute Classification

Semi-Supervised Semantic Segmentation using Redesigned Self-Training for White Blood Cells

MUSCLE: Multi-task Self-supervised Continual Learning to Pre-train Deep Models for X-ray Images of Multiple Body Parts

Self-NeRF: A Self-Training Pipeline for Few-Shot Neural Radiance Fields

Taming Contrast Maximization for Learning Sequential, Low-latency, Event-based Optical Flow

Explainable Action Prediction through Self-Supervision on Scene Graphs

Enhancing Self-Training Methods

Demonstrate-Search-Predict: Composing retrieval and language models for knowledge-intensive NLP

1Cademy @ Causal News Corpus 2022: Leveraging Self-Training in Causality Classification of Socio-Political Event Data

KD-MVS: Knowledge Distillation Based Self-supervised Learning for Multi-view Stereo

CELESTIAL: Classification Enabled via Labelless Embeddings with Self-supervised Telescope Image Analysis Learning

Multi-task Self-distillation for Graph-based Semi-Supervised Learning