A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Filters
Coded ResNeXt: a network for designing disentangled information paths
[article]
2022
Preserved Fulltext
Notably, leveraging coding theory to design the information paths enables us to use intermediate network layers for making early predictions without having to evaluate the full network. ...
We show that using our algorithm we can extract a lighter single-purpose binary classifier for a particular class by removing the parameters that do not participate in the predefined information path of ...
The work [47] bears resemblance to ours where they identify such information paths; a key difference is that they use a post hoc method so that the paths are identified and not designed. ...
doi:10.48550/arxiv.2202.05343
fatcat:36yrz7wrx5cgfovp6atzs2x76m
Stochastic Downsampling for Cost-Adjustable Inference and Improved Regularization in Convolutional Networks
2018
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
We propose a novel approach for cost-adjustable inference in CNNs -Stochastic Downsampling Point (SDPoint). ...
Sharing network parameters across different instances provides significant regularization boost. During inference, one may handpick a SDPoint instance that best fits the inference budget. ...
The ROSE Lab is supported by the National Research Foundation, Singapore, and the Infocomm Media Development Authority, Singapore. 2 It is a standard practice [10, 11, 40, 14] to resize images to a ...
doi:10.1109/cvpr.2018.00827
dblp:conf/cvpr/KuenK00YST18
fatcat:fmjdxcxt2fgjpn55kt3or5oeau
Frequency Disentangled Residual Network
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
In this paper, a frequency disentangled residual network (FDResNet) is proposed to tackle these issues. ...
Moreover, the present residual networks are not able to utilize the high and low frequency information suitably, which also challenges the generalization capability of the network. ...
We would like to acknowledge the NVIDIA for supporting with NVIDIA GPUs and Google Colab service for providing free computational resources which have been used in this research for the experiments. ...
arXiv:2109.12556v2
fatcat:hiu5boclhrgopb3bx4w7wzmv4a
Advancing Tassel Detection and Counting: Annotation and Algorithms
2021
Remote Sensing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
High-resolution RGB imagery acquired by unmanned aerial vehicles (UAVs), coupled with advanced machine learning approaches, including deep learning (DL), provides a new capability for monitoring flowering ...
Tassel counts provide valuable information related to flowering and yield prediction in maize, but are expensive and time-consuming to acquire via traditional manual approaches. ...
Changye Yang, Enyu Cai, and Jethro Zhou for their contributions to data annotation. ...
doi:10.3390/rs13152881
fatcat:fmp5ghb7zbhi5gjy5jviofeqwy
Variational Nested Dropout
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Based on this approach, we design a Bayesian nested neural network that learns the order knowledge of the parameter distributions. ...
For nested nets, when network parameters are removed, the performance decays in a human-specified trajectory rather than in a trajectory learned from data. ...
a single hyper-parameter β which helps to quantify the degree of learnt disentanglement. ...
arXiv:2101.11353v2
fatcat:wlugbqz5srej3jgrx22xthm3xi
Improved Residual Networks for Image and Video Recognition
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Our proposed improvements address all three main components of a ResNet: the flow of information through the network layers, the residual building block, and the projection shortcut. ...
In the deep learning era, we establish a new milestone for the depth of a CNN. ...
Our proposed approach facilitates learning by offering a better path for information to propagate through the network. ...
arXiv:2004.04989v1
fatcat:ghvk4fbr6nd7lf7rk345bbq5iq
Crossmodality Person Reidentification Based on Global and Local Alignment
2022
Wireless Communications and Mobile Computing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
In this paper, a novel AGF network is proposed for RGB-IR re-ID task, which is based on the idea of global and local alignment. ...
By automatically learning the importance of different channel features, it strengthens the ability of the network to extract more fine-grained structural information of person crossmodalities. ...
, which consists of two parts, one is a dual-path network for feature extraction, and the other is a bidirectional dual- constrained top-ranking loss for feature learning. ...
doi:10.1155/2022/4330804
fatcat:5f3v7wjk4nhd7hlhj4wr4o2rim
Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effect
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2020 pre-print arXiv:2009.12991v3 fatcat:dmup7jm4aresjhpc46o74pffyiOther Versions
We achieve new state-of-the-arts on three long-tailed visual recognition benchmarks: Long-tailed CIFAR-10/-100, ImageNet-LT for image classification and LVIS for instance segmentation. ...
Our framework elegantly disentangles the paradoxical effects of the momentum, by pursuing the direct causal effect caused by an input sample. ...
After equipped with ResNeXt-101-32x4d and ResNeXt-101-64x4d [2] for ImageNet-LT [9] and LVIS [7] V0.5, respectively, the proposed method gains additional improvements. ...
arXiv:2009.12991v4
fatcat:iirc3ar72bhelm7kjv2thcssxa
EfficientDet: Scalable and Efficient Object Detection
2020
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
In this paper, we systematically study neural network architecture design choices for object detection and propose several key optimizations to improve efficiency. ...
resolution, depth, and width for all backbone, feature network, and box/class prediction networks at the same time. ...
Acknowledgements Special thanks to Golnaz Ghiasi, Adams Yu, Daiyi Peng for their help on infrastructure and discussion. We also thank Adam Kraft, Barret Zoph, Ekin D. ...
doi:10.1109/cvpr42600.2020.01079
dblp:conf/cvpr/TanPL20
fatcat:nohfrx22jvb63n37irx65lroou
Virtual sawing using generative adversarial networks
2021
2021 36th International Conference on Image and Vision Computing New Zealand (IVCNZ)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
The specific aims were to choose an appropriate generative adversarial network architecture to build a trainable model as an extension to an existing virtual sawing system. ...
The main objective of this thesis was to study the suitability of state-of-the-art generative adversarial networks for virtual sawing. ...
The vanilla structure comprises a contracting path (left side) and an expansive path (right side) as shown in Figure 38 . ...
doi:10.1109/ivcnz54163.2021.9653436
fatcat:unb3q4fvmjdxzao37yrzyqo2am
EfficientDet: Scalable and Efficient Object Detection
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
In this paper, we systematically study neural network architecture design choices for object detection and propose several key optimizations to improve efficiency. ...
resolution, depth, and width for all backbone, feature network, and box/class prediction networks at the same time. ...
Acknowledgements Special thanks to Golnaz Ghiasi, Adams Yu, Daiyi Peng for their help on infrastructure and discussion. We also thank Adam Kraft, Barret Zoph, Ekin D. ...
arXiv:1911.09070v7
fatcat:nzgqoegmlvhk3hlkyomzhtsmfu
BlockDrop: Dynamic Inference Paths in Residual Networks
2018
2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
Preserved Fulltext
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.
In particular, given a pretrained ResNet, we train a policy network in an associative reinforcement learning setting for the dual reward of utilizing a minimal number of blocks while preserving recognition ...
Exploiting the robustness of Residual Networks (ResNets) to layer dropping, our framework selects on-the-fly which residual blocks to evaluate for a given novel image. ...
While this happens in standard ResNets as well, all images necessarily utilize all the paths, and disentangling this information is not possible. Instance difficulty. ...
doi:10.1109/cvpr.2018.00919
dblp:conf/cvpr/WuNKRDGF18
fatcat:b4vxf6d3mjgjro3kvf6rznsvoq
BlockDrop: Dynamic Inference Paths in Residual Networks
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
In particular, given a pretrained ResNet, we train a policy network in an associative reinforcement learning setting for the dual reward of utilizing a minimal number of blocks while preserving recognition ...
Exploiting the robustness of Residual Networks (ResNets) to layer dropping, our framework selects on-the-fly which residual blocks to evaluate for a given novel image. ...
While this happens in standard ResNets as well, all images necessarily utilize all the paths, and disentangling this information is not possible. Instance difficulty. ...
arXiv:1711.08393v4
fatcat:hrmjqu2p6raqlbhi65bx4z3x2u
CNN Variants for Computer Vision: History, Architecture, Application, Challenges and Future Scope
2021
Electronics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Significant emphasis has been given to leveraging channel and spatial information, with a depth of architecture and information processing via multi-path. ...
Furthermore, the introduction of large amounts of data and readily available hardware has opened new avenues for CNN study. ...
Acknowledgments: The authors would like to thank the reviewers for their valuable suggestions which helped in improving the quality of this paper. ...
doi:10.3390/electronics10202470
fatcat:aqhrysjtbjagzl6byalgy2du5a
Delving into Robust Object Detection from Unmanned Aerial Vehicles: A Deep Nuisance Disentanglement Approach
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
(NDFT), for the specific challenging problem of object detection in UAV images, achieving a substantial gain in robustness to those nuisances. ...
The code is available at https://github.com/TAMU-VITA/UAV-NDFT. ...
a specific time/location with each UAV flight's time-stamp and spatial location (or path). ...
arXiv:1908.03856v2
fatcat:d7gx37iz2nbrhiqg27axtl64nu
« Previous
Showing results 1 — 15 out of 49 results