Adobe Systems Inc

2007

Citations in the corpus

• Convolutional feature masking for joint object and stuff segmentation (2014, methodology): “We have further shown that convolutional feature masking The network is trained on the PASCAL-CONTEXT training set [18], and is applied on MIT-Adobe FiveK [3]. is applicable for joint object and stuff segmentation.”
• The Unreasonable Effectiveness of Deep Features as a Perceptual Metric (2018, methodology): “Our training and test sets contain patches sampled from the MIT5k [7] and RAISE1k [10] datasets, respectively (3) a test set containing real algorithm outputs, containing patches from a variety of applications.”
• Low-Light Image Enhancement via a Deep Hybrid Network (2019, methodology): “Since we focus on low-light image enhancement, we carefully select 336 input and output pairs that meet our requirements from the Adobe FiveK dataset [41].”
• Loss Functions for Image Restoration With Neural Networks (2017, methodology): “We trained the network considering different cost functions (`1, `2, SSIM5, SSSIM9, MS-SSIM and MSSSIM+`1)(4) on a training set of 700 RGB images taken from the MIT-Adobe FiveK Dataset [25], resized to a size of 999×666.”
• Learning a Deep Single Image Contrast Enhancer from Multi-Exposure Images (2018, unknown): “Our idea is inspired by the success of discriminative learning methods [18], [19], [20], especially the deep convolutional neural network (CNN) methods [21], [22], in image restoration.”
• Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement (2023, methodology): “We eveluate our method on LOL (v1 [54] and v2 [59]), SID [9], SMID [10], SDSD [48], and FiveK [5] datasets.”
• NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Images (2021, background): “Many works have tried to automate this process by us-ing heuristics or machine learning to map directly from raw data to postprocessed LDR images [7,9,17,22].”
• Generating Training Data for Denoising Real RGB Images via Camera Pipeline Simulation (2019, methodology): “We use the MIT-Adobe5k dataset [6] as our input images because it has high-quality photographs.”
• Toward Convolutional Blind Denoising of Real Photographs (2018, methodology): “1 to generate the synthetic noisy images, and use 400 images from BSD500 [40], 1600 images from Waterloo [36], and 1600 images from MIT-Adobe FiveK dataset [7] as the training data.”
• Learning Image-Adaptive 3D Lookup Tables for High Performance Photo Enhancement in Real-Time (2020, methodology): “The pio-neering work along this line is reported in [3], where Bychkovsky et al. developed the MIT-Adobe FiveK dataset and learned a mapping curve to approximate photographers’ adjustments.”
• CycleISP: Real Image Restoration via Improved Data Synthesis (2020, methodology): “To train the CycleISP model, we use the MIT-Adobe FiveK dataset [10], which contains 5000 RAW images.”
• Real Image Denoising With Feature Attention (2019, methodology): “To generate noisy synthetic images, we employ BSD500 [41], DIV2K [4], and MIT-Adobe FiveK [13], resulting in 4k images while for real noisy images, we use cropped patches of 512 × 512 from SSID [1], Poly [51], and RENOIR [5].”
• Exploring CLIP for Assessing the Look and Feel of Images (2022, methodology): “We then extend our scope to finegrained attributes using synthetic data and common restoration benchmarks (Wei et al. 2018; Bychkovsky et al. 2011; Xu et al. 2018; Rim et al. 2020).”
• Learned perceptual image enhancement (2017, methodology): “The enhancement CNN is trained and evaluated on the MIT-Adobe FiveK dataset [3].”
• IR-NAS: Neural Architecture Search for Image Restoration (2019, methodology): “Firstly, we use the camera pipeline simulation method to add noise to 25k patches extracted from MIT-Adobe5k dataset [33], then manually pick up 1812 patches which have the most realistic visual effects and finally randomly select 1600 patches as training set and reserve the rest as test set.”
• Ultra-High-Definition Low-Light Image Enhancement: A Benchmark and Transformer-Based Method (2022, methodology): “MIT-Adobe FiveK (Bychkovsky et al. 2011) consists of 5, 000 images captured of various indoor and outdoor scenes under different lighting conditions.”
• Beyond Brightening Low-light Images (2021, methodology): “For the deep learning methods, including Retinex-Net (Wei et al. 2018), GLAD (Wang et al. 2018), KinD (Zhang et al. 2019) and KinD++, we have retrained their models based on the 4500 pair images from the FiveK (Pic# 0001-4500).”
• Deep bilateral learning for real-time image enhancement (2017, methodology): “Besides HDR+ and the face brightening dataset, all the effects were applied to the unprocessed set of the MIT “FiveK” dataset [Bychkovsky et al. 2011].”
• Low-Light Image and Video Enhancement Using Deep Learning: A Survey (2021, methodology): “In addition to the testing subsets in the paired datasets [2], [4], [6], [8], [9], [55], [79], there are several testing data collected from related works or commonly used for experimental comparisons.”
• Fast End-to-End Trainable Guided Filter (2018, methodology): “Method L0 smoothing [39] Detail Manipulation [13] Style Transfer [3] Non-local Dehazing [5] Image Retouching [6] MSE PSNR SSIM MSE PSNR SSIM MSE PSNR SSIM MSE PSNR SSIM MSE PSNR SSIM”