AVA: A large-scale database for aesthetic visual analysis

By Naila Murray, Luca Marchesotti, Florent Perronnin. 2012

1,101 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Computer Vision and Pattern Recognition, Sensory Systems, Evaluation & Benchmarks

How researchers see this paper

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

In Cognitive Neuroscience

In Computer Vision and Pattern Recognition

In Sensory Systems

In Evaluation & Benchmarks

Citations in the corpus

Hierarchical Text-Conditional Image Generation with CLIP Latents (2022, methodology): “Next, we trained a CLIP linear probe to predict human aesthetic judgments using the AVA dataset [33] (Appendix A).”
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric (2018, background): “A related line of work is on No-Reference Image Quality Assessment (NR-IQA), such as AVA [38] and LIVE In the Wild [18].”
Recognizing Image Style (2013, methodology): “We also test on the AVA aesthetic prediction task [22], and show that using the “deep” object recognition features improves over the state-of-the-art results.”
NIMA: Neural Image Assessment (2017, methodology): “As discussed in [1], distribution of ratings with mean value between 2 and 8 can be closely approximated by Gaussian functions, and highly skewed ratings can be modeled by Gamma distributions.”
From Fidelity to Perceptual Quality: A Semi-Supervised Approach for Low-Light Image Enhancement (2020, methodology): “The high-quality images selected from aesthetic visual analysis dataset [21] based on the MOS values are used to represent the prior knowledge of human perception.”
Exploring Video Quality Assessment on User Generated Contents from Aesthetic and Technical Perspectives (2022, methodology): “For aesthetic rating , we select example images with good , fair and bad aesthetic quality from the aesthetic assessment database AVA [22], each for 20 images, as calibration for aesthetic evaluation.”
MaxViT: Multi-Axis Vision Transformer (2022, methodology): “We validated the efficacy of our proposed model on various vision tasks: ImageNet classification [46], image object detection and instance segmentation [50], image aesthetics/quality assessment [59], and unconditional image generation [23].”
Equal But Not The Same: Understanding the Implicit Relationship Between Persuasive Images and Text (2018, background): “age match in their speciﬁcity and concreteness. Image creativity and memorability. Images in the media are often more creative than typical photographs. Several works examine the aesthetics of photos [26,34] and the creativity of videos [32]. Some study the memorability of images [15,19]. In our wor”
From Patches to Pictures (PaQ-2-PiQ): Mapping the Perceptual Space of Picture Quality (2019, background): “Data collection began by sampling about 40 K highly diverse contents of diverse sizes and aspect ratios from hundreds of thousands of pictures drawn from public databases, including AVA [7], VOC [54], EMOTIC [55], and CERTH Blur [56].”
Human Preference Score: Better Aligning Text-to-image Models with Human Preference (2023, background): “The MLP is trained on several aesthetic datasets, including both real images and generated images (e.g., AVA [25], SAC [28]) to predict aesthetic scores ranging from 1 to 10.”
TOPIQ: A Top-Down Approach From Semantics to Distortions for Image Quality Assessment (2023, methodology): “For NR datasets, we have got CLIVE [77], KonIQ-10k [27], SPAQ [78], FLIVE [79] and AVA [28].”
A Deep Network Solution for Attention and Aesthetics Aware Photo Cropping (2019, methodology): “AA AVA [63] 2012 250,000 @ @ network”
Deep Classifiers from Image Tags in the Wild (2015, background): “Some of these missing tags are covered by scene [22, 30, 31] and style databases [11, 18].”
Photo Aesthetics Ranking Network with Attributes and Content Adaptation (2016, methodology): “For fair comparison, we follow the experimental practices and train/test split used in literature [16,17,23] which results in”