The power of comparative reasoning

Yagnik, Jay; Strelow, Dennis; Ross, David A.; Lin, Ruei-Sung

doi:10.1109/iccv.2011.6126527

Cited by 104 publications

(109 citation statements)

References 22 publications

Supporting

Mentioning

108

Contrasting

Unclassified

Order By: Relevance

“…While this computational complexity is high, there is room for optimism that our approach can be made significantly more efficient. For instance, the recent part-based hashing technique of [10] could be directly applied to our algorithm by applying winner-take-all hashing [40] on the discriminative visual elements. As we have a similar number of part detectors as considered in their work, we believe we can process an image in less than 20 seconds on a single multicore processor.…”

Section: Resultsmentioning

confidence: 99%

Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models

Aubry

Maturana

Efros

et al. 2014

2014 IEEE Conference on Computer Vision and Pattern Recognition

433

403

View full text Add to dashboard Cite

This paper poses object category detection in images as a type of 2D-to-3D alignment problem, utilizing the large quantities of 3D CAD models that have been made publicly available online. Using the "chair" class as a running example, we propose an exemplar-based 3D category representation, which can explicitly model chairs of different styles as well as the large variation in viewpoint. We develop an approach to establish part-based correspondences between 3D CAD models and real photographs. This is achieved by (i) representing each 3D model using a set of view-dependent mid-level visual elements learned from synthesized views in a discriminative fashion, (ii) carefully calibrating the individual element detectors on a common dataset of negative images, and (iii) matching visual elements to the test image allowing for small mutual deformations but preserving the viewpoint and style constraints. We demonstrate the ability of our system to align 3D models with 2D objects in the challenging PASCAL VOC images, which depict a wide variety of chairs in complex scenes.

show abstract

Section: Resultsmentioning

confidence: 99%

Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models

Aubry

Maturana

Efros

et al. 2014

2014 IEEE Conference on Computer Vision and Pattern Recognition

433

403

View full text Add to dashboard Cite

show abstract

“…Within the context of our classification scheme, for object classification one would want to deploy rotation variant features and additional, spatial constraints. Further research can try and robustify the image/patch descriptors by means of promising techniques such as WTA-hash, recently introduced in [38]. Also, we believe that starting from our classification framework and adding training at any level can improve performance further.…”

Section: Discussionmentioning

confidence: 99%

“…Recently, the power of comparative reasoning [38] has been shown and a Winner Take All (WTA) hash technique proposed. WTA-hash is a sparse embedding method transforming the input feature space into binary codes.…”

Section: Robustified Representation -(Wta-hash)mentioning

confidence: 99%

“…The power of comparative reasoning (WTA hash [38]) enhances and robustifies the representations by adding resilience to perturbations in numeric values.…”

Section: )mentioning

confidence: 99%

“…It combines the robustness and simplicity of local descriptors such as BIFs [9], layered, spatial information embedding similar to SPM [15] or a set of regions as in [13], the power of comparative reasoning [38], and state-of-the-art training-free classifiers [28]. The approach aims at addressing most of the problems encountered in texture classification, and exhibits the following advantages: 1) No need for training, and thus data independence.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Training-free Classification Framework for Textures, Writers, and Materials

Timofte¹,

Gool²

2012

Procedings of the British Machine Vision Conference 2012

View full text Add to dashboard Cite

We advocate the idea of a training-free texture classification scheme. This we demonstrate not only for traditional texture benchmarks, but also for the identification of materials and of the writers of musical scores. State-of-the-art methods operate using local descriptors, their intermediate representation over trained dictionaries, and classifiers. For the first two steps, we work with pooled local Gaussian derivative filters and a small dictionary not obtained through training, resp. Moreover, we build a multi-level representation similar to a spatial pyramid which captures region-level information. An extra step robustifies the final representation by means of comparative reasoning. As to the classification step, we achieve robust results using nearest neighbor classification, and state-of-the-art results with a collaborative strategy. Also these classifiers need no training.To the best of our knowledge, the proposed system yields top results on five standard benchmarks: 99.4% for CUReT, 97.3% for Brodatz, 99.5% for UMD, 99.4% for KTHTIPS, and 99% for UIUC. We significantly improve the state-of-the-art for three other benchmarks: KTHTIPS2b -66.3% (from 58.1%), CVC-MUSCIMA -99.8% (from 77.0%), and FMD -55.8% (from 54%).

show abstract

Kinome‐Wide Profiling Prediction of Small Molecules

Sorgenfrei

Merget

2017

ChemMedChem

View full text Add to dashboard Cite

Extensive kinase profiling data, covering more than half of the human kinome, are available nowadays and allow the construction of activity prediction models of high practical utility. Proteochemometric (PCM) approaches use compound and protein descriptors, which enables the extrapolation of bioactivity values to thus far unexplored kinases. In this study, the potential of PCM to make large-scale predictions on the entire kinome is explored, considering the applicability on novel compounds and kinases, including clinically relevant mutants. A rigorous validation indicates high predictive power on left-out kinases and superiority over individual kinase QSAR models for new compounds. Furthermore, external validation on clinically relevant mutant kinases reveals an excellent predictive power for mutations spread across the ATP binding site.

show abstract

The power of comparative reasoning

Cited by 104 publications

References 22 publications

Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models

Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models

A Training-free Classification Framework for Textures, Writers, and Materials

Kinome‐Wide Profiling Prediction of Small Molecules

Contact Info

Product

Resources

About