The Cube++ Illumination Estimation Dataset

Ershov, Egor; Savchik, Alexey; Semenkov, Illya; Banić, Nikola; Белокопытов, А. В.; Senshina, Daria; Koščević, Karlo; Subašić, Marko; Lončarić, Sven

doi:10.1109/access.2020.3045066

Cited by 19 publications

(12 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The network was trained on a proprietary dataset of annotated images, reaching 98.33% accuracy on an independent test set. Prior to being classified, the images in the AWB datasets were sRGB-rendered and whitebalanced in order to match the appearance expected by For reference, Figure 7 shows example images from the SimpleCube++ dataset 20 identified in terms of lighting conditions and illumination measure.…”

Section: Shooting Parameters and Illumination Levelsmentioning

confidence: 99%

Analysis of biases in automatic white balance datasets and methods

Buzzelli

Zini

Bianco

et al. 2022

Color Research & Application

View full text Add to dashboard Cite

Annotated datasets for automatic white balance (AWB) are used for the evaluation and, when necessary, the training, of AWB methods. Relying on such datasets requires awareness of the potential bias in their content and characteristics: some methods are designed to rely on the presence of particular elements, such as human skin, while other methods learn implicit relationships between image content and light properties from training data. The dependency on these relationships makes it fundamental to understand whether the available datasets are actually representative of common application scenarios, such as the presence of human subjects, the diversity of composition, or the illumination conditions. In this paper we overview the most common datasets for Automatic White Balance, including those for single as well as multiple illuminant estimation, providing a critical analysis on their characteristics. Furthermore, we identify a number of existing methods for single illuminant estimation, as a representative pool of approaches to the problem with various levels of complexity. We investigate how the performance of these correlate to the image content of common datasets.

show abstract

Section: Shooting Parameters and Illumination Levelsmentioning

confidence: 99%

Analysis of biases in automatic white balance datasets and methods

Buzzelli

Zini

Bianco

et al. 2022

Color Research & Application

View full text Add to dashboard Cite

show abstract

“…ILSVRC2012 [75] is a low resolution, JPG encoded subset of ImageNet [16] and is a popular and commonly acknowledged dataset for visual object recognition. Cube++ is a high-resolution dataset and comes in two flavors: as 16-bit encoded PNGs and JPGs [19].…”

Section: CVmentioning

confidence: 99%

Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines

Isenko,

Mayer,

Jedele

et al. 2022

Preprint

View full text Add to dashboard Cite

Preprocessing pipelines in deep learning aim to provide sufficient data throughput to keep the training processes busy. Maximizing resource utilization is becoming more challenging as the throughput of training processes increases with hardware innovations (e.g., faster GPUs, TPUs, and inter-connects) and advanced parallelization techniques that yield better scalability. At the same time, the amount of training data needed in order to train increasingly complex models is growing. As a consequence of this development, data preprocessing and provisioning are becoming a severe bottleneck in end-to-end deep learning pipelines.In this paper, we provide an in-depth analysis of data preprocessing pipelines from four different machine learning domains. We introduce a new perspective on efficiently preparing datasets for endto-end deep learning pipelines and extract individual trade-offs to optimize throughput, preprocessing time, and storage consumption. Additionally, we provide an open-source profiling library that can automatically decide on a suitable preprocessing strategy to maximize throughput. By applying our generated insights to real-world usecases, we obtain an increased throughput of 3× to 13× compared to an untuned system while keeping the pipeline functionally identical. These findings show the enormous potential of data pipeline tuning.

show abstract

“…We devide the NR2R dataset into two parts: 120 images are used for training and the rest 30 images are used for testing. The stage 1 was first pretrained on Cube++ dataset [12] which is captured within the same camera of the NR2R.Then we use Adam optimizer [23] with 5e-5 learning rate. Then the stage 2 was trained for 300 epochs with the same learning rate while the parameters of stage 1 was frozen.…”

Section: Training Detailsmentioning

confidence: 99%

Rendering Nighttime Image Via Cascaded Color and Brightness Compensation

Li¹,

Yi²,

Ma³

2022

Preprint

View full text Add to dashboard Cite

Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge [13] for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website 1 .

show abstract

The Cube++ Illumination Estimation Dataset

Cited by 19 publications

References 61 publications

Analysis of biases in automatic white balance datasets and methods

Analysis of biases in automatic white balance datasets and methods

Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines

Rendering Nighttime Image Via Cascaded Color and Brightness Compensation

Contact Info

Product

Resources

About