Hedging your bets: Optimizing accuracy-specificity trade-offs in large scale visual recognition

By Jia Deng, Jonathan Krause, Alexander C. Berg, Li Fei-Fei. 2012

220 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Optimization & Training

Citations in the corpus

Grafit: Learning fine-grained image representations with coarse labels (2020, background): “In computer vision, the concept of granularity underlies several tasks, such as fine-grained [13, 29] or hierarchical image classification [18, 60, 65].”
TreeTalk: Composition and Compression of Trees for Image Descriptions (2014, background): “Now that visual recognition algorithms are beginning to produce reliable estimates of image content (Perronnin et al., 2012; Deng et al., 2012a; Deng et al., 2010; Krizhevsky et al., 2012), the time seems ripe to begin exploring higher level semantic tasks.”
Question Relevance in VQA: Identifying Non-Visual And False-Premise Questions (2016, background): “A related idea is to avoid a highly speciﬁc prediction if there is a chance of being wrong, instead making a more generic prediction that is more likely to be right (Deng et al., 2012).”
HD-CNN: Hierarchical Deep Convolutional Neural Networks for Large Scale Visual Recognition (2014, methodology): “In [5], the predefined category hierarchy of the ImageNet dataset is utilized to achieve trade-offs between classification accuracy and specificity.”
Predicting Failures of Vision Systems (2014, unknown): “[10] makes”
Reasoning about Object Affordances in a Knowledge Base Representation (2014, background): “Object categories form a hierarchy consisting of several levels of abstraction [8].”
Learning More Universal Representations for Transfer-Learning (2017, unknown): “vision [32], [33], [34], [35].”
Dynamic-Structured Semantic Propagation Network (2018, methodology): “[8] used an accuracy-specificity trade-off algorithm to explore the WordNet hierarchy while Ordonez et al.”
Generalizing Image Captions for Image-Text Parallel Corpus (2013, background): “Note that some dependencies such as det ( ←→ ) would hold regardless of the particular 3 Code was provided by Deng et al. (2012).”
ImageNet Auto-Annotation with Segmentation Propagation (2014, background): “Several works tackle image classification [18,42,21,20,36, 17] or object detection in the fully supervised setting [17].”
Learning Semantic Image Representations at a Large Scale (2014, methodology): “Our method exhibits a good correlation with the human results, although it has a smaller generalization probability for L0 queries, possibly because current visual models are still not completely accurate in identifying leaf node classes [28].”
Incremental Learning of NCM Forests for Large-Scale Image Classification (2014, background): “It is also possible to exploit given or computed class hierarchies [8, 17, 19].”
Automating model search for large scale machine learning (2015, methodology): “The ﬁrst dataset is a pre-featurized version of the ImageNet 2010 dataset [12], featurized using a procedure attributed to [18].”
Open-vocabulary Object Retrieval (2014, methodology): “…cold tea, usually served in a glass with ice . . . popular packaged drink . . . ” Here the italicized words are terms that are also found in the corresponding synset descriptions of the object to which the user was referring, thus this projection expands the query to include category-level words.”
Semantic Aware Video Transcription Using Random Forest Classifiers (2014, background): “[8] observed the existence of a trade-off between accuracy and specificity for object categories and applied it to image classification.”
TuPAQ: An Efficient Planner for Large-scale Predictive Analytic Queries (2015, methodology): “The first dataset is a pre-featurized version of the ImageNet Large Scale Visual Recognition Challenge 2010 (ILSVRC2010) dataset [12], featurized using a procedure attributed to [19].”