Naive Bayes Super-Resolution Forest

Salvador, Jordi; Pérez-Pellitero, Eduardo

doi:10.1109/iccv.2015.45

Cited by 117 publications

(78 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For SISR, we study dictionary and deep learning methods. The dictionary methods are example-based ridge regression (EBSR) [9], sparse coding (ScSR) [77], Naive Bayes SR Interpolation BICUBIC NUISR [74], WNUISR [18] DBRSR [75] Non-blind L1BTV [22], BEPSR [78] reconstruction IRWSR [23] Blind reconstruction BVSR [24], SRB [62] forests (NBSRF) [10], and adjusted anchored neighborhood regression (A+) [12]. The deep learning methods comprise CNNs (SRCNN) [13], very deep networks (VDSR) [14], and deeply-recursive networks (DRCN) [15].…”

Section: Evaluated Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

Toward Bridging the Simulated-to-Real Gap: Benchmarking Super-Resolution on Real Data

Köhler

Bätz

Naderi

et al. 2019

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Capturing ground truth data to benchmark super-resolution (SR) is challenging. Therefore, current quantitative studies are mainly evaluated on simulated data artificially sampled from ground truth images. We argue that such evaluations overestimate the actual performance of SR methods compared to their behavior on real images. Toward bridging this simulated-to-real gap, we introduce the Super-Resolution Erlangen (SupER) database, the first comprehensive laboratory SR database of all-real acquisitions with pixel-wise ground truth. It consists of more than 80k images of 14 scenes combining different facets: CMOS sensor noise, real sampling at four resolution levels, nine scene motion types, two photometric conditions, and lossy video coding at five levels. As such, the database exceeds existing benchmarks by an order of magnitude in quality and quantity. This paper also benchmarks 19 popular single-image and multi-frame algorithms on our data. The benchmark comprises a quantitative study by exploiting ground truth data and qualitative evaluations in a large-scale observer study. We also rigorously investigate agreements between both evaluations from a statistical perspective. One interesting result is that top-performing methods on simulated data may be surpassed by others on real data. Our insights can spur further algorithm development, and the publicy available dataset can foster future evaluations.

show abstract

Section: Evaluated Algorithmsmentioning

confidence: 99%

“…In SISR, the use of external training data outperforms selfexemplar approaches [8]. Surprisingly, depending on the quality measure, popular deep nets [13], [14], [15] do not clearly outperform classical methods [9], [10], [12], [77]. This contrasts benchmarks on simulated data, where deep nets perform best.…”

Section: Remarks On the Quantitative Studymentioning

confidence: 99%

Toward Bridging the Simulated-to-Real Gap: Benchmarking Super-Resolution on Real Data

Köhler

Bätz

Naderi

et al. 2019

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…We differentiate between the two very different paradigms of prior-based single-image SR and redundancy-based multi-image SR. Prior-based Single-image SR. The goal of single-image SR is to fill in HR image patterns by leveraging a prior derived either from similar patches in other parts of the input image (self-examplars) [14,21], or from similar image patches from an existing image database [13], or -most commonly -from previously seen training data [25,48,53,54]. Recently the technologies of choice for learning the prior have been (deep) convolutional networks [10,64,46,22,66,49] and generative adversarial networks [31,60,42,61,59].…”

Section: Related Workmentioning

confidence: 99%

Learned Multi-View Texture Super-Resolution

Richard

Cherabier

Oswald

et al. 2019

2019 International Conference on 3D Vision (3DV)

View full text Add to dashboard Cite

We present a super-resolution method capable of creating a high-resolution texture map for a virtual 3D object from a set of lower-resolution images of that object. Our architecture unifies the concepts of (i) multi-view superresolution based on the redundancy of overlapping views and (ii) single-view super-resolution based on a learned prior of high-resolution (HR) image structure. The principle of multi-view super-resolution is to invert the image formation process and recover the latent HR texture from multiple lower-resolution projections. We map that inverse problem into a block of suitably designed neural network layers, and combine it with a standard encoder-decoder network for learned single-image super-resolution. Wiring the image formation model into the network avoids having to learn perspective mapping from textures to images, and elegantly handles a varying number of input views. Experiments demonstrate that the combination of multi-view observations and learned prior yields improved texture maps.

show abstract

“…Other advances within this family of algorithms include the Naive Bayes SR Forest [18], that uses an ensemble of bimodal trees in order to benefit from antipodal invariance. It is also highly competitive in speed thanks to the naive Bayes selection procedure which applies a single regressor per patch, differently from other forest approaches, e.g.…”

Section: B State Of the Artmentioning

confidence: 99%

Antipodally Invariant Metrics for Fast Regression-Based Super-Resolution

Pérez-Pellitero

Salvador

Ruiz-Hidalgo

et al. 2016

IEEE Trans. on Image Process.

Self Cite

View full text Add to dashboard Cite

Abstract-Dictionary-based Super-Resolution algorithms usually select dictionary atoms based on distance or similarity metrics. Although the optimal selection of nearest neighbors is of central importance for such methods, the impact of using proper metrics for Super-Resolution (SR) has been overlooked in literature, mainly due to the vast usage of Euclidean distance. In this paper we present a very fast regression-based algorithm which builds on densely populated anchored neighborhoods and sublinear search structures. We perform a study of the nature of the features commonly used for SR, observing that those features usually lie in the unitary hypersphere, where every point has a diametrically opposite one, i.e. its antipode, with same module and angle, but opposite direction. Even though we validate the benefits of using antipodally invariant metrics, most of the binary splits use Euclidean distance, which does not handle antipodes optimally. In order to benefit from both worlds, we propose a simple yet effective Antipodally Invariant Transform (AIT) that can be easily included in the Euclidean distance calculation. We modify the original Spherical Hashing algorithm with this metric in our Antipodally Invariant Spherical Hashing scheme, obtaining the same performance as a pure antipodally invariant metric. We round up our contributions with a novel feature transform that obtains a better coarse approximation of the input image thanks to Iterative Back Projection. The performance of our method, which we named Antipodally Invariant SuperResolution (AIS), improves quality (PSNR) and it is faster than any other state-of-the-art method.

show abstract

Naive Bayes Super-Resolution Forest

Cited by 117 publications

References 28 publications

Toward Bridging the Simulated-to-Real Gap: Benchmarking Super-Resolution on Real Data

Toward Bridging the Simulated-to-Real Gap: Benchmarking Super-Resolution on Real Data

Learned Multi-View Texture Super-Resolution

Antipodally Invariant Metrics for Fast Regression-Based Super-Resolution

Contact Info

Product

Resources

About