DeepFL-IQA: Weak Supervision for Deep IQA Feature Learning
This work addresses image quality assessment for artificially distorted images, which is important for applications like image processing and computer vision, but it is incremental as it builds on existing multi-level deep-feature approaches.
The authors tackled the problem of image quality assessment (IQA) for artificially distorted images by creating a large dataset (KADIS-700k with 700k images) and a weakly supervised feature learning method, achieving better performance than other feature-based no-reference IQA methods and matching the best existing deep learning methods on average across five benchmark databases.
Multi-level deep-features have been driving state-of-the-art methods for aesthetics and image quality assessment (IQA). However, most IQA benchmarks are comprised of artificially distorted images, for which features derived from ImageNet under-perform. We propose a new IQA dataset and a weakly supervised feature learning approach to train features more suitable for IQA of artificially distorted images. The dataset, KADIS-700k, is far more extensive than similar works, consisting of 140,000 pristine images, 25 distortions types, totaling 700k distorted versions. Our weakly supervised feature learning is designed as a multi-task learning type training, using eleven existing full-reference IQA metrics as proxies for differential mean opinion scores. We also introduce a benchmark database, KADID-10k, of artificially degraded images, each subjectively annotated by 30 crowd workers. We make use of our derived image feature vectors for (no-reference) image quality assessment by training and testing a shallow regression network on this database and five other benchmark IQA databases. Our method, termed DeepFL-IQA, performs better than other feature-based no-reference IQA methods and also better than all tested full-reference IQA methods on KADID-10k. For the other five benchmark IQA databases, DeepFL-IQA matches the performance of the best existing end-to-end deep learning-based methods on average.