Coreset Sampling from Open-Set for Fine-Grained Self-Supervised Learning
- URL: http://arxiv.org/abs/2303.11101v2
- Date: Fri, 24 Mar 2023 05:15:38 GMT
- Title: Coreset Sampling from Open-Set for Fine-Grained Self-Supervised Learning
- Authors: Sungnyun Kim, Sangmin Bae, Se-Young Yun
- Abstract summary: We introduce a novel Open-Set Self-Supervised Learning problem under the assumption that a large-scale unlabeled open-set is available.
In our problem setup, it is crucial to consider the distribution mismatch between the open-set and target dataset.
We demonstrate that SimCore significantly improves representation learning performance through extensive experimental settings.
- Score: 10.57079240576682
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: Deep learning in general domains has constantly been extended to
domain-specific tasks requiring the recognition of fine-grained
characteristics. However, real-world applications for fine-grained tasks suffer
from two challenges: a high reliance on expert knowledge for annotation and
necessity of a versatile model for various downstream tasks in a specific
domain (e.g., prediction of categories, bounding boxes, or pixel-wise
annotations). Fortunately, the recent self-supervised learning (SSL) is a
promising approach to pretrain a model without annotations, serving as an
effective initialization for any downstream tasks. Since SSL does not rely on
the presence of annotation, in general, it utilizes the large-scale unlabeled
dataset, referred to as an open-set. In this sense, we introduce a novel
Open-Set Self-Supervised Learning problem under the assumption that a
large-scale unlabeled open-set is available, as well as the fine-grained target
dataset, during a pretraining phase. In our problem setup, it is crucial to
consider the distribution mismatch between the open-set and target dataset.
Hence, we propose SimCore algorithm to sample a coreset, the subset of an
open-set that has a minimum distance to the target dataset in the latent space.
We demonstrate that SimCore significantly improves representation learning
performance through extensive experimental settings, including eleven
fine-grained datasets and seven open-sets in various downstream tasks.
Related papers
- Semi-Supervised One-Shot Imitation Learning [83.94646047695412]
One-shot Imitation Learning aims to imbue AI agents with the ability to learn a new task from a single demonstration.
We introduce the semi-supervised OSIL problem setting, where the learning agent is presented with a large dataset of trajectories.
We develop an algorithm specifically applicable to this semi-supervised OSIL setting.
arXiv Detail & Related papers (2024-08-09T18:11:26Z) - Few-Shot Learning for Annotation-Efficient Nucleus Instance Segmentation [50.407071700154674]
We propose to formulate annotation-efficient nucleus instance segmentation from the perspective of few-shot learning (FSL)
Our work was motivated by that, with the prosperity of computational pathology, an increasing number of fully-annotated datasets are publicly accessible.
Extensive experiments on a couple of publicly accessible datasets demonstrate that SGFSIS can outperform other annotation-efficient learning baselines.
arXiv Detail & Related papers (2024-02-26T03:49:18Z) - Knowledge Combination to Learn Rotated Detection Without Rotated
Annotation [53.439096583978504]
Rotated bounding boxes drastically reduce output ambiguity of elongated objects.
Despite the effectiveness, rotated detectors are not widely employed.
We propose a framework that allows the model to predict precise rotated boxes.
arXiv Detail & Related papers (2023-04-05T03:07:36Z) - M-Tuning: Prompt Tuning with Mitigated Label Bias in Open-Set Scenarios [103.6153593636399]
We propose a vision-language prompt tuning method with mitigated label bias (M-Tuning)
It introduces open words from the WordNet to extend the range of words forming the prompt texts from only closed-set label words to more, and thus prompts are tuned in a simulated open-set scenario.
Our method achieves the best performance on datasets with various scales, and extensive ablation studies also validate its effectiveness.
arXiv Detail & Related papers (2023-03-09T09:05:47Z) - Discrete Key-Value Bottleneck [95.61236311369821]
Deep neural networks perform well on classification tasks where data streams are i.i.d. and labeled data is abundant.
One powerful approach that has addressed this challenge involves pre-training of large encoders on volumes of readily available data, followed by task-specific tuning.
Given a new task, however, updating the weights of these encoders is challenging as a large number of weights needs to be fine-tuned, and as a result, they forget information about the previous tasks.
We propose a model architecture to address this issue, building upon a discrete bottleneck containing pairs of separate and learnable key-value codes.
arXiv Detail & Related papers (2022-07-22T17:52:30Z) - Continual Coarse-to-Fine Domain Adaptation in Semantic Segmentation [22.366638308792734]
Deep neural networks are typically trained in a single shot for a specific task and data distribution.
In real world settings both the task and the domain of application can change.
We introduce the novel task of coarse-to-fine learning of semantic segmentation architectures in presence of domain shift.
arXiv Detail & Related papers (2022-01-18T13:31:19Z) - Reconstruction guided Meta-learning for Few Shot Open Set Recognition [31.49168444631114]
We propose Reconstructing Exemplar-based Few-shot Open-set ClaSsifier (ReFOCS)
By using a novel exemplar reconstruction-based meta-learning strategy ReFOCS streamlines FSOSR.
We show ReFOCS to outperform multiple state-of-the-art methods.
arXiv Detail & Related papers (2021-07-31T23:23:35Z) - Hyperspherical embedding for novel class classification [1.5952956981784217]
We present a constraint-based approach applied to representations in the latent space under the normalized softmax loss.
We experimentally validate the proposed approach for the classification of unseen classes on different datasets using both metric learning and the normalized softmax loss.
Our results show that not only our proposed strategy can be efficiently trained on larger set of classes, as it does not require pairwise learning, but also present better classification results than the metric learning strategies.
arXiv Detail & Related papers (2021-02-05T15:42:13Z) - Learning to Count in the Crowd from Limited Labeled Data [109.2954525909007]
We focus on reducing the annotation efforts by learning to count in the crowd from limited number of labeled samples.
Specifically, we propose a Gaussian Process-based iterative learning mechanism that involves estimation of pseudo-ground truth for the unlabeled data.
arXiv Detail & Related papers (2020-07-07T04:17:01Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.