A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf
.
Filters
Towards Optimal Structured CNN Pruning via Generative Adversarial Learning
2019
2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
We then effectively solve the optimization problem by generative adversarial learning (GAL), which learns a sparse soft mask in a label-free and an end-to-end manner. ...
Besides, these methods are designed for pruning a specific structure, such as filter or block structures without jointly pruning heterogeneous structures. ...
We propose a generative adversarial learning (GAL) to effectively conduct structured pruning of CNNs. ...
doi:10.1109/cvpr.2019.00290
dblp:conf/cvpr/LinJYZCYHD19
fatcat:hxlpqa2dlvf2hf2cgs56anjaba
Towards Optimal Structured CNN Pruning via Generative Adversarial Learning
[article]
2019
arXiv
pre-print
We then effectively solve the optimization problem by generative adversarial learning (GAL), which learns a sparse soft mask in a label-free and an end-to-end manner. ...
Besides, these methods are designed for pruning a specific structure, such as filter or block structures without jointly pruning heterogeneous structures. ...
We propose a generative adversarial learning (GAL) to effectively conduct structured pruning of CNNs. ...
arXiv:1903.09291v1
fatcat:wltbctcrojes5mwjpkfezhnjfm
Towards Higher Ranks via Adversarial Weight Pruning
[article]
2023
arXiv
pre-print
This rank-based optimization objective guides sparse weights towards a high-rank topology. ...
However, unstructured pruning presents a structured pattern at high pruning rates, which limits its performance. ...
We also gratefully thank Yuxin Zhang and Xiaolong Ma for their generous help. ...
arXiv:2311.17493v1
fatcat:42hk7ecqe5dzxo43c57kvrxmd4
VCIP 2020 Index
2020
2020 IEEE International Conference on Visual Communications and Image Processing (VCIP)
Stereoscopic image reflection removal based o
Wasserstein Generative Adversarial Network
Luo, Jixiang
Spatial-Channel Context-Based Entropy
Modeling for End-to-end Optimized Image
Compression ...
Adversarial Networks
Han, Xiao
A Dense-Gated U-Net for Brain Lesion
Segmentation
Han, Xiao
Text-to-Image Generation via Semi-Supervised
Training
Han, Xiao
Low Resolution Facial Manipulation ...
doi:10.1109/vcip49819.2020.9301896
fatcat:bdh7cuvstzgrbaztnahjdp5s5y
A Hybrid Bayesian-Convolutional Neural Network for Adversarial Robustness
2022
IEICE transactions on information and systems
We keep the remainder of CNNs unchanged. We adopt the Bayes without Bayesian Learning (BwoBL) algorithm for hyBCNN networks to execute Bayesian inference towards adversarial robustness. ...
transfer learning. ...
We focus on controlling a pair of structural hyperparameters of BC and BA layers to generate the stochastic models and execute Bayesian inference towards adversarial robustness. ...
doi:10.1587/transinf.2021edp7239
fatcat:2my24bznlncjxcdafviklfvyle
Channel Pruning via Automatic Structure Search
[article]
2020
arXiv
pre-print
And then, we formulate the search of optimal pruned structure as an optimization problem and integrate the ABC algorithm to solve it in an automatic manner to lessen human interference. ...
of pruned structure can be significantly reduced. ...
., 2019] proposed a sparsity-regularized mask for channel pruning, which is optimized through a data-driven selection or generative adversarial learning. ...
arXiv:2001.08565v3
fatcat:unl2vpbikvb4vjz7iwdlbe63uy
Deep-Learning Steganalysis for Removing Document Images on the Basis of Geometric Median Pruning
2020
Symmetry
redundant filters as extensively as possible through the overall iterative pruning and artificial bee colony (ABC) automatic pruning algorithms to reduce the size of the network structure of the existing ...
While the steganography secret message is primarily removed via active steganalysis. ...
[39] proposed a sparse regularization mask method based on channel pruning; the mask is optimized via data-driven selection or generative adversarial learning. Zhao et al. ...
doi:10.3390/sym12091426
fatcat:5s5pakblmbbuzoraepgvcqj4zq
GAN-Knowledge Distillation for One-stage Object Detection
2020
IEEE Access
The feature maps generated by teacher network and student network are employed as true and fake samples respectively, and generating adversarial training for both of them to improve the performance of ...
Convolutional neural networks (CNN) have a significant improvement in the accuracy of object detection. ...
Ensemble via Adversarial Learning), [33] proposed by Xu et al., and [34] proposed by Liu et al. ...
doi:10.1109/access.2020.2983174
fatcat:nw6fvq5qtrcjzhjfcez6zz66ta
Fingerprinting Multi-exit Deep Neural Network Models via Inference Time
[article]
2021
arXiv
pre-print
under comprehensive adversarial settings. ...
In this paper, we propose a novel approach to fingerprint multi-exit models via inference time rather than inference predictions. ...
To transform a CNN into an SDN, we add and optimize the ICs with Adam optimizer of learning rate 0.001 decayed by 0.1 at epoch 15. ...
arXiv:2110.03175v1
fatcat:c23dl2w4jvfovoogqebcujfpxy
VeriCompress: A Tool to Streamline the Synthesis of Verified Robust Compressed Neural Networks from Scratch
[article]
2023
arXiv
pre-print
When deployed on a resource-restricted generic platform, these models require 5-8 times less memory and 2-4 times less inference time than models used in verified robustness literature. ...
PM inherits parameter weights from a pre-trained dense model, and the PM structure is learned through global magnitude-based pruning. ...
DST's higher efficacy than the static-maskbased sparse-training is attributed to high gradient flow allowing the model to learn an optimal sparse network for effective inference generation (Evci et al ...
arXiv:2211.09945v7
fatcat:gxw2kmne3rdibmdmskjeavc7re
Recent Advances in Understanding Adversarial Robustness of Deep Neural Networks
[article]
2020
arXiv
pre-print
We give preliminary definitions on what adversarial attacks and robustness are. After that, we study frequently-used benchmarks and mention theoretically-proved bounds for adversarial robustness. ...
Adversarial examples are inevitable on the road of pervasive applications of deep neural networks (DNN). ...
CNN models are instead strongly biased towards learning the representation of textures rather than shapes. ...
arXiv:2011.01539v1
fatcat:e3o47epftbc2rebpdx5yotzriy
Overcoming Long-term Catastrophic Forgetting through Adversarial Neural Pruning and Synaptic Consolidation
[article]
2021
IEEE Transactions on Neural Networks and Learning Systems
accepted
a structure-aware parameter-importance measurement and an element-wise parameter updating strategy, decreases the cumulative error when learning new tasks. ...
Inspired by the memory consolidation mechanism in mammalian brains with synaptic plasticity, we propose a confrontation mechanism in which Adversarial Neural Pruning and synaptic Consolidation (ANPyC) ...
This finding proves that ANPSC has strong generalization performance in MLP, CNN and VAE.
5) Continual learning in GAN: We further apply the ANPSC to a generative adversarial network [37] . ...
doi:10.1109/tnnls.2021.3056201
pmid:33577459
arXiv:1912.09091v2
fatcat:glic2itroraa7jpicjaamjljsu
Channel Pruning via Automatic Structure Search
2020
Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence
And then, we formulate the search of optimal pruned structure as an optimization problem and integrate the ABC algorithm to solve it in an automatic manner to lessen human interference. ...
of pruned structure can be significantly reduced. ...
., 2019] proposed a sparsity-regularized mask for channel pruning, which is optimized through a data-driven selection or generative adversarial learning. ...
doi:10.24963/ijcai.2020/94
dblp:conf/ijcai/LinJZZW020
fatcat:kxiizioxbzethmjxqd53jl6xkq
Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safety
[article]
2021
arXiv
pre-print
This work provides a structured and broad overview of them. ...
These shortcomings are diverse and range from a lack of generalization over insufficient interpretability to problems with malicious inputs. ...
Furthermore, this research has been funded by the Federal Ministry of Education and Research of Germany as part of the competence center for machine learning ML2R (01IS18038B). ...
arXiv:2104.14235v1
fatcat:f6sj3v2brza7thyzw7b7fkpo2m
Machine Learning for Microcontroller-Class Hardware – A Review
[article]
2022
arXiv
pre-print
This paper highlights the unique requirements of enabling onboard machine learning for microcontroller class devices. ...
The advancements in machine learning opened a new opportunity to bring intelligence to the low-end Internet-of-Things nodes such as microcontrollers. ...
Some frameworks [51] [54] provide support for structured pruning, allowing policies for channel and filter pruning rather than pruning weights in an irregular fashion. ...
arXiv:2205.14550v3
fatcat:y272riitirhwfgfiotlwv5i7nu
« Previous
Showing results 1 — 15 out of 1,354 results