-
Notifications
You must be signed in to change notification settings - Fork 2
/
cldd24.bib
58 lines (58 loc) · 12 KB
/
cldd24.bib
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
% if the proceedings is a reissued proceedings, please add the field
% 'firstpublished' to the entry below, giving the original data of
% publication in YYYY-MM-DD format.
@proceedings{CLD2-2024,
year = 2024,
booktitle = {Proceedings of The Workshop on Classifier Learning from Difficult Data},
address = {Santiago de Compostella, Spain},
volume = 999999,
name = {Classifier Learning from Difficult Data},
shortname = {CLD2},
editor = {Zyblewski, Pawel and Grana, Manuel and Ksieniewicz Pawel and Minku, Leandro},
start = {2024-10-19},
end = {2024-10-20},
published = {2024-10-21},
conference_url = {https://cldd.kssk.pwr.edu.pl}
}
@inproceedings{trajdos24,
title = {A dual ensemble classifier used to recognise contaminated multi-channel EMG and MMG signals in the control of upper limb bioprosthesis},
author = {Trajdos, Pawel and Kurzynski, Marek},
pages = {1--8},
abstract = {Myopotential pattern recognition to decode the intent of the user is the most advanced approach to controlling a powered bioprosthesis. Unfortunately, many factors make this a difficult problem and achieving acceptable recognition quality in real-word conditions is a serious challenge. The aim of the paper is to develop a recognition system that will mitigate factors related to multimodality and multichannel recording of biosignals and their high susceptibility to contamination. The proposed method involves the use of two co-operating multiclassifier systems. The first system is composed of one-class classifiers related to individual electromyographic (EMG) and mechanomyographic (MMG) biosignal recording channels, and its task is to recognise contaminated channels. The role of the second system is to recognise the class of movement resulting from the patient's intention. The ensemble system consists of base classifiers using the representation (extracted features) of biosignals from different channels. The system uses a dynamic selection mechanism, eliminating those base classifiers that are associated with biosignal channels that are recognised by the one-class ensemble system as being contaminated. Experimental studies were conducted using signals from an able-bodied person with simulation of amputation. The results obtained confirm the hypothesis that the use of a dual ensemble classifier leads to improved classification quality.}
}
@inproceedings{seo24,
title = {Keypoint Mask-based Local Feature Matching and Keypoint Erasing-based Global Feature Representations for Visible-Infrared Person Re-Identification},
author = {Seo, Kisung and Gwon, Soonyong and Chae Woon},
pages = {9--16},
abstract = {In Visible-Infrared Person Re-Identification, the most crucial challenge is reducing the Modality discrepancy between visible and infrared images. To address this, various approaches such as data augmentation, generative transformation for the opposite modality, and extraction of Modality-Shared and Modality-Specific features have been attempted. While each approach has contributed significantly to performance improvement, re-identification remains particularly challenging when dealing with individuals who have similar clothing or body shapes but are different persons. It is mainly due to the inconsistency in representing identical local features across cross-modalities in existing methods. This paper proposes a novel learning representations of keypoint-based local and global features. Keypoint-based masking for local feature representation learning aims to normalize the representations of each keypoint's locality, thereby reducing the modality gap at the feature level. Representation learning for local features using keypoint-based masking reduces feature-level modality gaps by identifying the local representation of each keypoint. Representation learning for global features using keypoint-based eraing increases the efficiency and diversity of the global representation by generating images that cover the same area. We compare our proposed methodology and various existing methods for the mAP and Rank-1 performances on the SYSU-MM01 datasets. Experimental results demonstrate that our proposed model effectively solves the existing key and critical problems.}
}
@inproceedings{wojtulewicz24,
title = {On Speeding Up the Training of Deep Neural Networks Using the Streaming Approach: The Base-Values Mechanism},
author = {Wojtulewicz, Mateusz and Duda, Piotr and Nowicki, Robert and Rutkowski, Leszek},
pages = {17--24},
abstract = {Efficient and stable neural network training is crucial for advancing machine learning applications. This study explores the promising streaming approach as an alternative to traditional epoch-based methods. This paradigm shift involves transforming training data into a continuous stream, prioritizing challenging examples to enhance the learning process. Building upon this approach, we introduce an innovative Base-Values mechanism aimed at further improving the speed and stability of the streaming training process. We apply this framework to the original streaming training algorithms, resulting in algorithms such as Persistent Loss-Based (PLB) and Persistent Entropy-Based algorithm (PEB). We conduct a comprehensive experimental comparison on EMNIST dataset, analyzing traditional epoch-based methods and the streaming approach, including both original and new methods employing the Base-Values mechanism. The exploration of various hyperparameters, including pre-training length, mini-batch size, and learning rate, provides valuable insights into their impact on the performance and stability of those neural network training methods. The results demonstrate the superior performance of a novel streaming algorithm that incorporates the Base-Values mechanism compared to both a traditional epoch-based method and other methods.}
}
@inproceedings{wojciechowski24,
title = {$F_\Beta$-plot - a visual tool for evaluationg imabalnced data classifiers},
author = {Wojciechowski, Szymon and Wozniak, Michal},
pages = {25--31},
abstract = {Imbalanced data classification suffers from a lack of reliable metrics. This runs primarily from the fact that for most real-life (and commonly used benchmark) problems, we do not have information from the user on the actual form of the loss function that should be minimized. Although it is pretty common to have metrics indicating the classification quality within each class, for the end user, the analysis of several such metrics is then required, which in practice causes difficulty in interpreting the usefulness of a given classifier. Hence, many aggregate metrics have been proposed or adopted for the imbalanced data classification problem, but there is still no consensus on which should be used. An additional disadvantage is their ambiguity and systematic bias toward one class. Moreover, their use in analyzing experimental results in recognition of those classification models that perform well for the chosen aggregated metrics is burdened with the abovementioned drawbacks. Hence, the paper proposes a simple approach to analyzing the popular parametric metric Fᵦ. We point out that it is possible to indicate for a given pool of analyzed classifiers when a given model should be preferred depending on user requirements.}
}
@inproceedings{pinitas24,
title = {Silhouette Distance Loss for Learning Few-Shot Contrastive Representations},
author = {Pinitas Kosmas and Rasajski, Nemanja and Mankantasis, Konstantinas and Yannakakis Georgios},
pages = {32--39},
abstract = {Conventional supervised contrastive learning methods excel in optimising encoders for discriminative tasks. In scenarios where only a few labelled samples are available, however, they struggle in eliminating the inductive bias when transferring from source to target classes. This is a byproduct (and inherent limitation) of their underlying optimisation process that involves training a representation to maximise class separation, without directly optimising for within-class cohesion. As a response to this limitation this paper introduces the Silhouette Distance (SD) loss, a new optimisation objective for supervised contrastive representation learning. SD aims to enhance the quality of learned embeddings by emphasising both the cohesion and separation of representation clusters for each class. We test SD extensively across several few-shot learning scenarios---where labelled data is limited---and we compare its performance against supervised contrastive loss and prototypical network loss for various text and image classification tasks. We also test SD in a cross-domain manner, by training a model on one dataset and testing it on another, within the same modality. Our results demonstrate the superior, at worst competitive, performance of the SD loss compared to its baselines. By leveraging pre-trained models and fine-tuning techniques, our study highlights how the SD loss can effectively improve representation learning across different modalities and domains. This initial study showcases the potential of the SD loss as a robust alternative within the few-shot learning setting.}
}
@inproceedings{hergert24,
title = {Detecting Noisy Labels Using Early Stopped Models},
author = {Hergert, Lea and Jelasity, Mark},
pages = {40--47},
abstract = {We are concerned with the problem of identifying samples with noisy labels in a given dataset. Using the predictions of a well-generalizing model to flag incorrectly predicted labels as noisy is a known method but it is not considered competitive. At the same time, it has been observed recently that gradient descent fits clean samples first, and the noisy samples are memorized later. Inspired by related theoretical results, we revisit the idea of using the predictions of an early stopped model to classify samples as noisy or clean. We offer two key improvements that allow this strikingly simple approach to outperform some well-known methods. First, we use the model over its own training set to directly exploit the so-called clean priority learning phenomenon. Second, we use an ensemble of model check points around the early stopping point to reduce the variance of the predictions. We also introduce a novel method that makes use of the same early stopped model ensemble, but classifies samples based on the per-sample gradient of the loss, motivated by recent theoretical results on clean priority learning. Our approaches only passively observe a normal training run and collect checkpoints. No extra input samples are added, no thresholds are tuned, and no pre-trained models are used. Through empirical evaluations, we demonstrate that our methods are competitive with other approaches from related work for both detecting noisy samples and for noise-filtered training.}
}
@inproceedings{daram24,
title = {Does Alignment help continual learning?},
author = {Daram, Anurag and Kudithipudi, Dhireesha},
pages = {48--55},
abstract = {Backpropagation relies on instantaneous weight transport and global updates, thus questioning its neural plausibility. Continual learning mechanisms that are largely biologically inspired employ backpropagation as the baseline training rule. In this work, we examine the role of learning rules that avoid the weight transport problem in the context of continual learning. We investigate weight estimation approaches that use linear combinations of local and non-local regularization primitives for alignment-based learning. We couple these approaches with parameter regularization and replay mechanisms to demonstrate robust continual learning capabilities. We show that the layer-wise operations observed in alignment-based learning help to boost performance. We evaluated the proposed models in complex task-aware and task-free scenarios on multiple image classification datasets. We study the dynamics of representational similarity for learning rules compared to backpropagation. Lastly, we provide a mapping of the representational similarity to the knowledge preservation capabilities of the models.}
}