Early eXit

Early exit (EE) techniques aim to improve the efficiency of deep neural networks by allowing predictions to be made at intermediate layers, reducing computational cost and latency. Current research focuses on developing adaptive algorithms and training strategies for EE models, often incorporating techniques like knowledge distillation, multi-armed bandits, and prototypical networks within architectures such as BERT and ResNet. This work is significant because it addresses the growing need for faster and more resource-efficient deep learning inference, particularly in resource-constrained environments like edge devices and mobile applications. The resulting improvements in speed and energy efficiency have broad implications across various fields, including image captioning, speech recognition, and natural language processing.

Papers

August 31, 2023

Dynamic nsNet2: Efficient Deep Noise Suppression with Early Exiting
Riccardo Miccini, Alaa Zniber, Clément Laroche, Tobias Piechowiak, Martin Schoeberl, Luca Pezzarossa, Ouassim Karrakchou, Jens Sparsø, Mounir Ghogho
Deep Learning Deep Architecture Early eXit Deep Noise Suppression Early Exiting

August 22, 2023

Using Early Exits for Fast Inference in Automatic Modulation Classification
Elsayed Mohammed, Omar Mashaal, Hatem Abou-Zeid
Deep Learning Deep Neural Network Fast Inference Deep Architecture Early eXit Synthetic Signal Automatic Modulation Classification

May 21, 2023

F-PABEE: Flexible-patience-based Early Exiting for Single-label and Multi-label text Classification Tasks
Xiangxiang Gao, Wei Zhu, Jiasheng Gao, Congrui Yin
Language Model Multi Label Classification Fast Inference Single Label Multi Label Text Classification Early eXit Early Exit

April 20, 2023

FIANCEE: Faster Inference of Adversarial Networks via Conditional Early Exits
Polina Karpikova, Radionova Ekaterina, Anastasia Yaschenko, Andrei Spiridonov, Leonid Kostyushko, Riccardo Fabbricatore, Aleksei Ivakhnenko
Adversarial Learning Image Synthesis Image Quality Human Relationship Fast Inference Generative Neural Network Early eXit

March 28, 2023

Robust and IP-Protecting Vertical Federated Learning against Unexpected Quitting of Parties
Jingwei Sun, Zhixu Du, Anna Dai, Saleh Baghersalimi, Alireza Amirshahi, David Atienza, Yiran Chen
Robust Version Structured Dropout Vertical Federated Learning Inference Attack Early eXit Political Party Passive Party

March 4, 2023

Hierarchical Training of Deep Neural Networks Using Early Exiting
Yamin Sepehri, Pedram Pad, Ahmet Caner Yüzügüler, Pascal Frossard, L. Andrea Dunbar
Deep Neural Network Edge Cloud Tiny ImageNet Early eXit Hierarchical Training

February 3, 2023

Enhancing Once-For-All: A Study on Parallel Blocks, Skip Connections and Early Exits
Simone Sarti, Eugenio Lomurno, Andrea Falanti, Matteo Matteucci
Neural Network Neural Architecture Search Study Feature Tiny ImageNet Early eXit Fit All Approach Honeycomb Block

January 28, 2023

Anticipate, Ensemble and Prune: Improving Convolutional Neural Networks via Aggregated Early Exits
Simone Sarti, Eugenio Lomurno, Matteo Matteucci
Neural Network Convolutional Neural Network Image Classification Multi Layer Diverse Ensemble Pruning Method Early eXit Aggregated Data

September 20, 2022

Unsupervised Early Exit in DNNs with Multiple Exits
Hari Narayan N U, Manjesh K. Hanawal, Avinash Bhardwaj
Deep Neural Network Early eXit Early Exit Multi Exit

September 17, 2022

Learning to Weight Samples for Dynamic Early-exiting Networks
Yizeng Han, Yifan Pu, Zihang Lai, Chaofei Wang, Shiji Song, Junfen Cao, Wenhui Huang, Chao Deng, Gao Huang
LeArning Abstract Inference Efficiency Early eXit Multi Exit Early Exit Network Weight Prediction

July 14, 2022

T-RECX: Tiny-Resource Efficient Convolutional neural networks with early-eXit
Nikhil P Ghanathe, Steve Wilton
Convolutional Neural Network TinyML Model Early eXit Small Sized Convolutional Neural Network

June 15, 2022

Resource-Constrained Edge AI with Early Exit Prediction
Rongkang Dong, Yuyi Mao, Jun Zhang
Early eXit Edge Artificial Intelligence Early Exit Network Device Edge Co Inference

June 9, 2022

Predictive Exit: Prediction of Fine-Grained Early Exits for Computation- and Energy-Efficient Inference
Xiangjie Li, Chenfei Lou, Zhengping Zhu, Yuchi Chen, Yingtao Shen, Yehan Ma, An Zou
Human Prediction Early eXit Early Exit Energy Efficient Inference Energy Efficient Deep Learning