RAIL-KD: RAndom intermediate layer mapping for knowledge distillation

MA Haidar, N Anchuri, M Rezagholizadeh… - arXiv preprint arXiv …, 2021 - arxiv.org
Intermediate layer knowledge distillation (KD) can improve the standard KD technique
(which only targets the output of teacher and student models) especially over large pre …

Boosting text augmentation via hybrid instance filtering framework

H Yang, K Li - Findings of the Association for Computational …, 2023 - aclanthology.org
Text augmentation is an effective technique for addressing the problem of insufficient data in
natural language processing. However, existing text augmentation methods tend to focus on …

On-the-fly denoising for data augmentation in natural language understanding

T Fang, W Zhou, F Liu, H Zhang, Y Song… - arXiv preprint arXiv …, 2022 - arxiv.org
Data Augmentation (DA) is frequently used to automatically provide additional training data
without extra human annotation. However, data augmentation may introduce noisy data that …

Image and Text: Fighting the Same Battle? Super-resolution Learning for Imbalanced Text Classification

R Meunier, F Benamara, V Moriceau… - 2023 Conference on …, 2023 - hal.science
In this paper, we propose SRL4NLP, a new approach for data augmentation by drawing an
analogy between image and text processing: Super-resolution learning. This method is …

Intended Target Identification for Anomia Patients with Gradient-based Selective Augmentation

J Kim, R Storaï, S Hwang - Findings of the Association for …, 2024 - aclanthology.org
In this study, we investigate the potential of language models (LMs) in aiding patients
experiencing anomia, a difficulty identifying the names of items. Identifying the intended …

Taming Prompt-Based Data Augmentation for Long-Tailed Extreme Multi-Label Text Classification

P Xu, M Song, Z Li, S Lu, L Jing… - ICASSP 2024-2024 IEEE …, 2024 - ieeexplore.ieee.org
In extreme multi-label text classification (XMC), labels usually follow a long-tailed
distribution, where most labels only contain a small number of documents and limit the …

CILDA: Contrastive data augmentation using intermediate layer knowledge distillation

MA Haidar, M Rezagholizadeh, A Ghaddar… - arXiv preprint arXiv …, 2022 - arxiv.org
Knowledge distillation (KD) is an efficient framework for compressing large-scale pre-trained
language models. Recent years have seen a surge of research aiming to improve KD by …

Learning Semantic Textual Similarity via Multi-Teacher Knowledge Distillation: A Multiple Data Augmentation method

Z Lu, Y Zhao, J Li, Y Tian - 2024 9th International Conference …, 2024 - ieeexplore.ieee.org
Data augmentation technologies, which can overcome the expensive and time-consuming
issue of high-quality labeled data generation in semantic textual similarity (STS) tasks, use …

A Simple Structure for Building a Robust Model

X Tan, J Gao, R Li - International Conference on Intelligence Science, 2022 - Springer
As deep learning applications, especially programs of computer vision, are increasingly
deployed in our lives, we have to think more urgently about the security of these …

Dehusked Coconut Vision-Based Counting on a Manufacturing Plant Utilizing the YOLOv8, ByteTrack, and Roboflow Algorithms

NJ Suganob, R Concepcion II, R Billones… - … Conference on Intelligent …, 2024 - Springer
Many manufacturing industries in third-world countries are still in need of process systems
improvement in order to increase productivity. Some manufacturing plants are encountering …