Deep Feature Space: A Geometrical Perspective

@article{Kansizoglou2020DeepFS,
  title={Deep Feature Space: A Geometrical Perspective},
  author={Ioannis Kansizoglou and Loukas Bampis and Antonios Gasteratos},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year={2020},
  volume={44},
  pages={6823-6838}
}
One of the most prominent attributes of Neural Networks (NNs) constitutes their capability of learning to extract robust and descriptive features from high dimensional data, like images. Hence, such an ability renders their exploitation as feature extractors particularly frequent in an abundance of modern reasoning systems. Their application scope mainly includes complex cascade tasks, like multi-modal recognition and deep Reinforcement Learning (RL). However, NNs induce implicit biases that… 
[PDF] Semantic Reader
12 Citations
Highly Influential Citations
1
Background Citations
6
Methods Citations
4
Results Citations
1

12 Citations

A Geometric Perspective towards Neural Calibration via Sensitivity Decomposition

Inspired by the decomposition, a simple extension to current softmax-linear models, which learns to disentangle the two components during training, outperforms other calibration methods on standard calibration metrics in the face of out-of-distribution data and corruption with significantly less complexity.

The Effect of Negative Samples on the Accuracy of Water Body Extraction Using Deep Learning Networks

It was found that when the training dataset contains a certain proportion of cloud shadow samples, the trained network can handle the misclassification of cloud shadows well and more accurately extract water bodies.

Tackling Dataset Bias With an Automated Collection of Real-World Samples

This study presents a novel pipeline that combines state-of-the-art modules to automatically create new thematic datasets with low bias and was able to acquire and allocate more than 880K previously unseen images to produce a data collection.

Continuous Emotion Recognition for Long-Term Behavior Modeling through Recurrent Neural Networks

This work introduces a novel approach that gradually maps and learns the personality of a human, by conceiving and tracking the individual's emotional variations throughout their interaction, and proposes a handy tool for HRI scenarios, where robot’s activity adaptation is needed for enhanced interaction performance and safety.

The Revisiting Problem in Simultaneous Localization and Mapping: A Survey on Visual Loop Closure Detection

This article describes a loop closure detection system’s structure, covering an extensive collection of topics, including the feature extraction, the environment representation, the decision-making step, and the evaluation process, and discusses open and new research challenges.

Rethinking maximum-margin softmax for adversarial robustness

Do Neural Network Weights account for Classes Centers?

A specific symmetry is proposed and studied both analytically and empirically that satisfies the above assumption, addressing the established convergence issues and is proven to ensure a more stable learning curve compared against the corresponding ones succeeded by popular models in the field of feature learning.

Web acquired image datasets need curation: an examplar pipeline evaluated on Greek food images

This paper suggests a pipeline consisting of state-of-the-art and off- the-shelf methods for curating an image dataset retrieved from the Web, aiming at increasing the accuracy in food recognition applications.

An Interpolation-Based Polynomial Method of Estimating the Objective Function Value in Scheduling Problems of Minimizing the Maximum Lateness

An approach to estimating the objective function value of minimization maximum lateness problem is proposed, it is shown how to use transformed instances to define a new continuous objective function, and two new polynomial cases are proposed.

Real-Time Monocular Human Depth Estimation and Segmentation on Embedded Systems

A novel, low complexity network architecture for fast and accurate human depth estimation and segmentation in indoor environments, aiming to applications for resource-constrained platforms with a monocular camera being the primary perception module.

References

SHOWING 1-10 OF 63 REFERENCES

L2-constrained Softmax Loss for Discriminative Face Verification

This paper adds an L2-constraint to the feature descriptors which restricts them to lie on a hypersphere of a fixed radius and shows that integrating this simple step in the training pipeline significantly boosts the performance of face verification.

Large-Margin Softmax Loss for Convolutional Neural Networks

A generalized large-margin softmax (L-Softmax) loss which explicitly encourages intra-class compactness and inter-class separability between learned features and which not only can adjust the desired margin but also can avoid overfitting is proposed.

Recognizing Human Emotional State From Audiovisual Signals

A novel multiclassifier scheme is proposed to boost the recognition performance of human emotional state from audiovisual signals based on a comparative study of different classification algorithms and specific characteristics of individual emotion.

Learning Multiple Layers of Features from Tiny Images

It is shown how to train a multi-layer generative model that learns to extract meaningful features which resemble those found in the human visual cortex, using a novel parallelization algorithm to distribute the work among multiple machines connected on a network.

BAUM-1: A Spontaneous Audio-Visual Face Database of Affective and Mental States

Baseline experimental results on the BAUM-1 database show that recognition of affective and mental states under naturalistic conditions is quite challenging, and the database is expected to enable further research on audio-visual affect and mental state recognition under close-to-real scenarios.

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

This paper designs a novel type of neural network that directly consumes point clouds, which well respects the permutation invariance of points in the input and provides a unified architecture for applications ranging from object classification, part segmentation, to scene semantic parsing.

A scalable active framework for region annotation in 3D shape collections

This work proposes a novel active learning method capable of enriching massive geometric datasets with accurate semantic region annotations, and demonstrates that incorporating verification of all produced labelings within this unified objective improves both accuracy and efficiency of the active learning procedure.

ImageNet: A large-scale hierarchical image database

A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.

On-line deep learning method for action recognition

The engagement of ART and the Viterbi Algorithm in a Deep learning architecture, here, for the first time, leads to a substantially different approach for action recognition, which is shown to outperform other deep learning methodologies, in terms of classification accuracy.

An Active Learning Paradigm for Online Audio-Visual Emotion Recognition

A novel paradigm for online emotion classification is provided, which exploits both audio and visual modalities and produces a responsive prediction when the system is confident enough, and is compared against other state-of-the-art models.
...