Filters








3,373 Hits in 3.1 sec

X-Class: Text Classification with Extremely Weak Supervision [article]

Zihan Wang and Dheeraj Mekala and Jingbo Shang
2020 arXiv   pre-print
Extensive experiments demonstrate that X-Class can rival and even outperform seed-driven weakly supervised methods on 7 benchmark datasets.  ...  ., based on topics and locations), so document representations must be adaptive to the given class names. We propose a novel framework X-Class to realize it.  ...  Compared Methods We compare with two seed-driven weakly supervised methods.  ... 
arXiv:2010.12794v1 fatcat:idgugnmqlffufbpjj7iph6qhb4

Not All Frames Are Equal: Weakly-Supervised Video Grounding With Contextual Similarity and Visual Clustering Losses

Jing Shi, Jia Xu, Boqing Gong, Chenliang Xu
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
We investigate the problem of weakly-supervised video grounding, where only video-level sentences are provided.  ...  This is a challenging task, and previous Multi-Instance Learning (MIL) based image grounding methods turn to fail in the video domain.  ...  In this work, we follow the proposal-based MIL methods [11, 40] due to the simplicity and effectiveness of the MIL learning framework. Weakly-supervised object localization.  ... 
doi:10.1109/cvpr.2019.01069 dblp:conf/cvpr/0005XGX19 fatcat:jbfo7ovqbbhctlk6nmminia5y4

LoGAN: Latent Graph Co-Attention Network for Weakly-Supervised Video Moment Retrieval [article]

Reuben Tan, Huijuan Xu, Kate Saenko, Bryan A. Plummer
2020 arXiv   pre-print
Prior strongly- and weakly-supervised approaches often leverage co-attention mechanisms to learn visual-semantic representations for localization.  ...  The goal of weakly-supervised video moment retrieval is to localize the video segment most relevant to the given natural language query without access to temporal annotations during training.  ...  It leverages the complementary nature of video-language pairs through a multi-level coattention mechanism to learn contextualized visual-semantic representations.  ... 
arXiv:1909.13784v2 fatcat:btgosisk6bb4pklnwgpkojk53m

Contrastive Learning for Weakly Supervised Phrase Grounding [article]

Tanmay Gupta, Arash Vahdat, Gal Chechik, Xiaodong Yang, Jan Kautz, Derek Hoiem
2020 arXiv   pre-print
A key idea is to construct effective negative captions for learning through language model guided word substitutions.  ...  Our weakly supervised phrase grounding model trained on COCO-Captions shows a healthy gain of 5.7% to achieve 76.7% accuracy on Flickr30K Entities benchmark.  ...  Weakly Supervised Phrase Grounding.  ... 
arXiv:2006.09920v3 fatcat:2fecqopa3jdjlf5psmjykdj52q

Weakly-supervised Contextualization of Knowledge Graph Facts

Nikos Voskarides, Edgar Meij, Ridho Reinanda, Abhinav Khaitan, Miles Osborne, Giorgio Stefanoni, Prabhanjan Kambadur, Maarten de Rijke
2018 The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval - SIGIR '18  
NFCM first generates a set of candidate facts in the neighborhood of a given fact and then ranks the candidate facts using a supervised learning to rank model.  ...  In order to obtain the annotations required to train the learning to rank model at scale, we generate training data automatically using distant supervision on a large entity-tagged text corpus.  ...  NFCM then ranks the candidate facts by how relevant they are for contextualizing the main fact. We estimate our learning to rank model using supervised data.  ... 
doi:10.1145/3209978.3210031 dblp:conf/sigir/VoskaridesMRKOS18 fatcat:kmcewxfyazgizdmn6fhdtibxle

Weakly Supervised Human-Object Interaction Detection in Video via Contrastive Spatiotemporal Regions [article]

Shuang Li, Yilun Du, Antonio Torralba, Josef Sivic, Bryan Russell
2021 arXiv   pre-print
We introduce the task of weakly supervised learning for detecting human and object interactions in videos.  ...  We demonstrate improved performance over weakly supervised baselines adapted to our task on our video dataset.  ...  Training objective In addition to the weakly supervised contrastive loss L ST , we propose a sparsity loss L spa , and a classification loss L cls for weakly supervised learning.  ... 
arXiv:2110.03562v1 fatcat:bdco4w4jcrdrnn2wpzabxaemgi

WELDON: Weakly Supervised Learning of Deep Convolutional Neural Networks

Thibaut Durand, Nicolas Thome, Matthieu Cord
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In this paper, we introduce a novel framework for WEakly supervised Learning of Deep cOnvolutional neural Networks (WELDON).  ...  Firstly, WELDON leverages recent improvements on the Multiple Instance Learning paradigm, i.e. negative evidence scoring and top instance selection.  ...  Another option to gain strong invariance is to explicitly align image regions, e.g. by using Weakly Supervised Learning (WSL) models.  ... 
doi:10.1109/cvpr.2016.513 dblp:conf/cvpr/DurandTC16 fatcat:2mqnc6l7jzhmhgjr55vispq4he

Weakly Supervised Affordance Detection

Johann Sawatzky, Abhilash Srikantha, Juergen Gall
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Our approach achieves a higher affordance detection accuracy than other weakly supervised methods that also rely on keypoint annotations or image annotations as weak supervision.  ...  In [26] a ranked weighted F-measure was proposed for measuring the accuracy for affordance detection.  ...  Recently, several weakly supervised approaches have been proposed for weakly supervised learning of image segmentations. An approach based on a CNN has been proposed in [27] .  ... 
doi:10.1109/cvpr.2017.552 dblp:conf/cvpr/SawatzkySG17 fatcat:d5f6jkkiljdjpaeofokspbbayq

ConceptLearner: Discovering visual concepts from weakly labeled image collections

Bolei Zhou, Vignesh Jagadeesh, Robinson Piramuthu
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
ConceptLearner shows promising performance compared to fully supervised and weakly supervised methods.  ...  Under domainspecific supervision, we further evaluate the learned concepts for scene recognition on SUN database and for object detection on Pascal VOC 2007.  ...  weakly supervised methods.  ... 
doi:10.1109/cvpr.2015.7298756 dblp:conf/cvpr/ZhouJP15 fatcat:hucw3uctg5eajdkk5v2bufsgdq

Seed Word Selection for Weakly-Supervised Text Classification with Unsupervised Error Estimation [article]

Yiping Jin, Akshay Bhatia, Dittaya Wanvarie
2021 arXiv   pre-print
Furthermore, in the weakly-supervised learning setting, we do not have any labeled document to measure the seed words' efficacy, making the seed word selection process "a walk in the dark".  ...  Weakly-supervised text classification aims to induce text classifiers from only a few user-provided seed words. The vast majority of previous work assumes high-quality seed words are given.  ...  Recently, disambiguate the seed words by explicitly learning different senses of each word with contextualized word embeddings.  ... 
arXiv:2104.09765v1 fatcat:of3w45ct55bvvjzgh6rytrnqgu

Neural Check-Worthiness Ranking with Weak Supervision: Finding Sentences for Fact-Checking

Casper Hansen, Christian Hansen, Stephen Alstrup, Jakob Grue Simonsen, Christina Lioma
2019 Companion Proceedings of The 2019 World Wide Web Conference on - WWW '19  
Our model is an end-to-end trainable neural network for checkworthiness ranking, which is trained on large amounts of unlabelled data through weak supervision.  ...  Motivated by this, we present a neural check-worthiness sentence ranking model that represents each word in a sentence by both its embedding (aiming to capture its semantics) and its syntactic dependencies  ...  Our model effectively incorporates weak supervision: using an existing check-worthiness ranking system to weakly label political speeches significantly improved performance.  ... 
doi:10.1145/3308560.3316736 dblp:conf/www/HansenHASL19 fatcat:w2yux2pqevetfndwyyiisxqnva

Weakly Supervised Graph Propagation Towards Collective Image Parsing

Si Liu, Shuicheng Yan, Tianzhu Zhang, Changsheng Xu, Jing Liu, Hanqing Lu
2012 IEEE transactions on multimedia  
In this work, we propose a weakly supervised graph propagation method to automatically assign the annotated labels at image level to those contextually derived semantic regions.  ...  Index Terms-Concept map-based image retrieval, convex concave programming (CCCP), image annotation, nonnegative multiplicative updating, weakly supervised image parsing.  ...  It transforms multiple instance problem into an input for a graph-based single instance semi-supervised learning method.  ... 
doi:10.1109/tmm.2011.2174780 fatcat:6i3fk4vfrjho3o7s3uw6w43joa

ConceptLearner: Discovering Visual Concepts from Weakly Labeled Image Collections [article]

Bolei Zhou, Vignesh Jagadeesh, Robinson Piramuthu
2014 arXiv   pre-print
ConceptLearner shows promising performance compared to fully supervised and weakly supervised methods.  ...  Under domain-specific supervision, we further evaluate the learned concepts for scene recognition on SUN database and for object detection on Pascal VOC 2007.  ...  For the concepts learned from NUS-WIDE dataset in Figure 5 are more contextual words ranked from tf-idf scores associated with the central concept name as the sub-category concept name.  ... 
arXiv:1411.5328v1 fatcat:yccr3e2k7baszfd2ttcz6rz2dm

Weak Supervision and Referring Attention for Temporal-Textual Association Learning [article]

Zhiyuan Fang, Shu Kong, Zhe Wang, Charless Fowlkes, Yezhou Yang
2020 arXiv   pre-print
We validate our WSRA through extensive experiments for temporally grounding by languages, demonstrating that it outperforms the state-of-the-art weakly-supervised methods notably.  ...  Therefore we provide a Weak-Supervised alternative with our proposed Referring Attention mechanism to learn temporal-textual association (dubbed WSRA).  ...  We measure the performance using the Rank@1 (R@1) and Rank@5 (R@5) (accuracy of the top-1/5 retrieved candidates) and their mean Intersection of Unions (mIoU) when IoU=1.  ... 
arXiv:2006.11747v2 fatcat:bpqa6chthfgjhatmsgqq5t2dym

Neural Check-Worthiness Ranking with Weak Supervision: Finding Sentences for Fact-Checking [article]

Casper Hansen, Christian Hansen, Stephen Alstrup, Jakob Grue Simonsen, Christina Lioma
2019 arXiv   pre-print
Our model is an end-to-end trainable neural network for check-worthiness ranking, which is trained on large amounts of unlabelled data through weak supervision.  ...  Motivated by this, we present a neural check-worthiness sentence ranking model that represents each word in a sentence by both its embedding (aiming to capture its semantics) and its syntactic dependencies  ...  Our model effectively incorporates weak supervision: using an existing check-worthiness ranking system to weakly label political speeches significantly improved performance.  ... 
arXiv:1903.08404v1 fatcat:7t2mihjvvff2fj2dtzhqrxddda
« Previous Showing results 1 — 15 out of 3,373 results