Discriminative Sample Aware Loss Function

Discriminative sample-aware loss functions aim to improve the performance of machine learning models by focusing on the relationships between data samples and their class labels. Current research explores their application across various tasks, including few-shot learning, recommender systems, and person re-identification, often integrating them with existing architectures like convolutional neural networks and transformers. These loss functions enhance model discriminability by explicitly considering intra- and inter-class variances, leading to improved classification accuracy, uncertainty estimation, and robustness to out-of-distribution data, ultimately advancing the capabilities of various machine learning applications.

Papers

March 7, 2024

Discriminative Sample-Guided and Parameter-Efficient Feature Space Adaptation for Cross-Domain Few-Shot Learning
Rashindrie Perera, Saman Halgamuge
LeArning Abstract Cross Domain Unseen Domain Discriminative Reply Cross Domain Few Shot Learning Parameter Efficient Adaptation Discriminative Sample Aware Loss Function

October 6, 2023

Making Users Indistinguishable: Attribute-wise Unlearning in Recommender Systems
Yuyuan Li, Chaochao Chen, Xiaolin Zheng, Yizhao Zhang, Zhongxuan Han, Dan Meng, Jun Wang
Recommender System User Base Recommendation Performance Knowledge Unlearning Level Unlearning Discriminative Sample Aware Loss Function Recommendation Unlearning

December 28, 2022

Joint Discriminative and Metric Embedding Learning for Person Re-Identification
Sinan Sabri, Zaigham Randhawa, Gianfranco Doretto
LeArning Abstract Person Re Identification North Star Metric Discriminative Task Metric Learning Loss Softmax Loss Discriminative Sample Aware Loss Function

May 12, 2022

Distinction Maximization Loss: Efficiently Improving Out-of-Distribution Detection and Uncertainty Estimation by Replacing the Loss and Calibrating
David Macêdo, Cleber Zanchettin, Teresa Ludermir
Distribution Detection Uncertainty Estimation Cross Entropy Loss Interest Loss Deterministic Neural Network Discriminative Sample Aware Loss Function

April 7, 2022

Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators
Yu Meng, Chenyan Xiong, Payal Bajaj, Saurabh Tiwary, Paul Bennett, Jiawei Han, Xia Song
Language Model Adversarial Learning Text Encoder Pre Training Model Pre Trained Text Adversarial Curriculum Training Data Generator Discriminative Sample Aware Loss Function

February 14, 2022

Discriminability-enforcing loss to improve representation learning
Florinel-Alin Croitoru, Diana-Nicoleta Grigore, Radu Tudor Ionescu
Convolutional Neural Network Deep Neural Network Representation Learning Discriminative Feature Level Representation Discriminative Power Discriminative Sample Aware Loss Function

Discriminative Sample Aware Loss Function

Papers

Discriminative Sample-Guided and Parameter-Efficient Feature Space Adaptation for Cross-Domain Few-Shot Learning

Making Users Indistinguishable: Attribute-wise Unlearning in Recommender Systems

Joint Discriminative and Metric Embedding Learning for Person Re-Identification

Distinction Maximization Loss: Efficiently Improving Out-of-Distribution Detection and Uncertainty Estimation by Replacing the Loss and Calibrating

Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators

Discriminability-enforcing loss to improve representation learning