Meta-learning & Few-shot LearningMeta learning is a branch of metacognition concerned with learning about one's own learning and learning processes.
While many novel meta-learning methods are proposed and some recent work argue that a pre-trained classifier is good enough for few-shot learning, our Meta-Baseline takes the strengths of both classification pre-training and meta-learning
Cited by13BibtexViews107Links
0
0
Yunxiao Qin, Chenxu Zhao, Xiangyu Zhu, Zezheng Wang, Zitong Yu, Tianyu Fu, Feng Zhou, Jingping Shi,Zhen Lei
AAAI, pp.11916-11923, (2020)
We develop a novel method Adaptive Inner-update Meta Face Anti-Spoofing and propose three zero- and few-shot face anti-spoofing benchmarks
Cited by7BibtexViews23Links
0
0
ICLR, (2020)
We show that meta-regularization in model-agnostic meta-learning can be rigorously motivated by a PAC-Bayes bound on generalization
Cited by4BibtexViews241Links
0
0
AAAI, pp.12087-12094, (2020)
Our method provides a new alternative to deal with K-way, N -shot few-shot segmentation, which is more meaningful than the conventional setting
Cited by4BibtexViews68Links
0
0
ICLR, (2020)
T-Neural Architecture Search learns a meta-architecture that is able to adapt to a new task and quickly through a few gradient steps, which is more flexible than the existing Neural Architecture Search methods
Cited by3BibtexViews114Links
0
0
Kye Seong Min, Lee Hae Beom, Kim Hoirin,Hwang Sung Ju
We proposed a novel transductive inference scheme for metric-based few-shot learning models
Cited by3BibtexViews6Links
0
0
Reddy Mahesh Kumar Krishna, Hossain Mohammad,Rochan Mrigank,Wang Yang
WACV, pp.2803-2812, (2020)
The key reason for this surge in interest is the demand of automated complex crowd scene understanding that appears in computer vision applications such as surveillance, traffic monitoring, etc
Cited by2BibtexViews16Links
0
0
Fei Nanyi,Lu Zhiwu, Gao Yizhao, Tian Jia,Xiang Tao,Wen Ji-Rong
Extensive experiments demonstrate that both meta-domain adaptation and meta-knowledge distillation significantly boost the performance of a variety of few-shot learning methods, resulting in new state-of-the-art on three benchmarks
Cited by1BibtexViews70Links
0
0
Shumin Deng, Ningyu Zhang, Zhanlin Sun,Jiaoyan Chen,Huajun Chen
AAAI, no. 10 (2020): 13773-13774
Results show that our model outperforms conventional state-of-the-art few-shot text classification models
Cited by0BibtexViews16Links
0
0
NeurIPS 2020, (2020)
We presented a novel casual framework: Interventional Few-Shot Learning, to address an overlooked deficiency in recent FSL methods: the pre-training is a confounder hurting the performance
Cited by0BibtexViews20Links
0
0
Trapit Bansal, Rishikesh Jha,Tsendsuren Munkhdalai, Andrew McCallum
EMNLP 2020, (2020)
This paper proposes a self-supervised approach to generate a large, rich, meta-learning task distribution from unlabeled text
Cited by0BibtexViews12Links
0
0
Few-shot natural language processing (NLP) refers to NLP tasks that are accompanied with merely a handful of labeled examples. This is a real-world challenge that an AI system must learn to handle. Usually we rely on collecting more auxiliary information or developing a more ef...
Cited by0BibtexViews5Links
0
0
IJCAI, pp.1090-1096, (2020)
By analyzing the characteristics of fine-grained images, we propose a multi-attention meta-learning method, which uses attention mechanisms of the base learner and task learner to capture discriminative parts of images according to the current task
Cited by0BibtexViews17Links
0
0
Alireza Rahimpour,Hairong Qi
WACV, pp.3168-3176, (2020)
We introduced a simple but effective fewshot learning model which can produce highly discriminative embedding space using the combination of proposed query and support set feature manipulation frameworks
Cited by0BibtexViews8Links
0
0
Vinay Kumar Verma, Dhanajit Brahma,Piyush Rai
national conference on artificial intelligence, (2020)
We have proposed a novel framework for Zero-Shot Learning and generalized ZSL which is based on the meta-learning framework over a conditional generative model
Cited by0BibtexViews10Links
0
0
Jyoti Narwariya,Pankaj Malhotra,Lovekesh Vig,Gautam Shroff, Vishnu TV
COMAD/CODS, pp.28-36, (2020)
Our approach Few-Shot UTSC Approach-2 with a simpler update rule than Few-Shot UTSC Approach-1 is the second best model but is very closely followed by the residual network models trained from scratch
Cited by0BibtexViews32Links
0
0
international conference on learning representations, (2019)
Gradient-based meta-learning techniques are both widely applicable and proficient at solving challenging few-shot learning and fast adaptation problems. However, they have the practical difficulties of operating in high-dimensional parameter spaces in extreme low-data regimes. We...
Cited by299BibtexViews189Links
0
0
CVPR, (2019): 10657-10665
We presented a meta-learning approach with convex base learners for few-shot learning
Cited by194BibtexViews66Links
0
0
computer vision and pattern recognition, (2019)
We show that our novel meta-transfer learning trained with hard task meta-batch learning curriculum achieves the top performance for tackling few-shot learning problems
Cited by169BibtexViews117Links
0
0
ICLR, (2019)
We propose Transductive Propagation Network, a novel meta-learning framework for transductive inference that classifies the entire test set at once to alleviate the low-data problem
Cited by157BibtexViews50Links
0
0
Keywords
Image ClassificationMeta LearningNeural NetworksCommon KnowledgeDecision BoundaryFast AdaptationFederated LearningFirst-order ApproximationHessian MatrixHierarchical Bayes
Authors
Sergey Levine
Paper 10
Chelsea Finn
Paper 9
Qianru Sun
Paper 5
Richard  Zemel
Paper 4
Yuxiong Wang
Paper 4
Bharath Hariharan
Paper 4
Tat-Seng Chua
Paper 4
Yoshua Bengio
Paper 4
Trevor Darrell
Paper 4
Bernt Schiele
Paper 4