Despite the impressive representation capacity of vision transformer models, current light-weight vision transformer models still suffer from inconsistent and incorrect dense predictions at local regions. We suspect that the power of their self-attention mechanism is limited in shallower and thinner networks. We propose Lite Vision Transformer (LVT), a novel light-weight transformer network with two enhanced self-attention mechanisms to improve the model performances for mobile deployment. For the low-level features, we introduce Convolutional Self-Attention (CSA). Unlike previous approaches of merging convolution and self-attention, CSA introduces local self-attention into the convolution within a kernel of size 3x3 to enrich low-level features in the first stage of LVT. For the high-level features, we propose Recursive Atrous Self-Attention (RASA), which utilizes the multi-scale context when calculating the similarity map and a recursive mechanism to increase the representation capability with marginal extra parameter cost. The superiority of LVT is demonstrated on ImageNet recognition, ADE20K semantic segmentation, and COCO panoptic segmentation. The code is made publicly available.
We consider the problem of obtaining image quality representations in a self-supervised manner. We use prediction of distortion type and degree as an auxiliary task to learn features from an unlabeled image dataset containing a mixture of synthetic and realistic distortions. We then train a deep Convolutional Neural Network (CNN) using a contrastive pairwise objective to solve the auxiliary problem. We refer to the proposed training framework and resulting deep IQA model as the CONTRastive Image QUality Evaluator (CONTRIQUE). During evaluation, the CNN weights are frozen and a linear regressor maps the learned representations to quality scores in a No-Reference (NR) setting. We show through extensive experiments that CONTRIQUE achieves competitive performance when compared to state-of-the-art NR image quality models, even without any additional fine-tuning of the CNN backbone. The learned representations are highly robust and generalize well across images afflicted by either synthetic or authentic distortions. Our results suggest that powerful quality representations with perceptual relevance can be obtained without requiring large labeled subjective image quality datasets. The implementations used in this paper are available at \url{https://github.com/pavancm/CONTRIQUE}.
The popularity of streaming videos with live, high-action content has led to an increased interest in High Frame Rate (HFR) videos. In this work we address the problem of frame rate dependent Video Quality Assessment (VQA) when the videos to be compared have different frame rate and compression factor. The current VQA models such as VMAF have superior correlation with perceptual judgments when videos to be compared have same frame rates and contain conventional distortions such as compression, scaling etc. However this framework requires additional pre-processing step when videos with different frame rates need to be compared, which can potentially limit its overall performance. Recently, Generalized Entropic Difference (GREED) VQA model was proposed to account for artifacts that arise due to changes in frame rate, and showed superior performance on the LIVE-YT-HFR database which contains frame rate dependent artifacts such as judder, strobing etc. In this paper we propose a simple extension, where the features from VMAF and GREED are fused in order to exploit the advantages of both models. We show through various experiments that the proposed fusion framework results in more efficient features for predicting frame rate dependent video quality. We also evaluate the fused feature set on standard non-HFR VQA databases and obtain superior performance than both GREED and VMAF, indicating the combined feature set captures complimentary perceptual quality information.
Image harmonization aims to improve the quality of image compositing by matching the "appearance" (\eg, color tone, brightness and contrast) between foreground and background images. However, collecting large-scale annotated datasets for this task requires complex professional retouching. Instead, we propose a novel Self-Supervised Harmonization framework (SSH) that can be trained using just "free" natural images without being edited. We reformulate the image harmonization problem from a representation fusion perspective, which separately processes the foreground and background examples, to address the background occlusion issue. This framework design allows for a dual data augmentation method, where diverse [foreground, background, pseudo GT] triplets can be generated by cropping an image with perturbations using 3D color lookup tables (LUTs). In addition, we build a real-world harmonization dataset as carefully created by expert users, for evaluation and benchmarking purposes. Our results show that the proposed self-supervised method outperforms previous state-of-the-art methods in terms of reference metrics, visual quality, and subject user study. Code and dataset are available at \url{https://github.com/VITA-Group/SSHarmonization}.
Image quality assessment (IQA) is an important research topic for understanding and improving visual experience. The current state-of-the-art IQA methods are based on convolutional neural networks (CNNs). The performance of CNN-based models is often compromised by the fixed shape constraint in batch training. To accommodate this, the input images are usually resized and cropped to a fixed shape, causing image quality degradation. To address this, we design a multi-scale image quality Transformer (MUSIQ) to process native resolution images with varying sizes and aspect ratios. With a multi-scale image representation, our proposed method can capture image quality at different granularities. Furthermore, a novel hash-based 2D spatial embedding and a scale embedding is proposed to support the positional embedding in the multi-scale representation. Experimental results verify that our method can achieve state-of-the-art performance on multiple large scale IQA datasets such as PaQ-2-PiQ, SPAQ and KonIQ-10k.
In this paper, we propose a generic model transfer scheme to make Convlutional Neural Networks (CNNs) interpretable, while maintaining their high classification accuracy. We achieve this by building a differentiable decision forest on top of CNNs, which enjoys two characteristics: 1) During training, the tree hierarchies of the forest are learned in a top-down manner under the guidance from the category semantics embedded in the pre-trained CNN weights; 2) During inference, a single decision tree is dynamically selected from the forest for each input sample, enabling the transferred model to make sequential decisions corresponding to the attributes shared by semantically-similar categories, rather than directly performing flat classification. We name the transferred model deep Dynamic Sequential Decision Forest (dDSDF). Experimental results show that dDSDF not only achieves higher classification accuracy than its conuterpart, i.e., the original CNN, but has much better interpretability, as qualitatively it has plausible hierarchies and quantitatively it leads to more precise saliency maps.
We develop an approach to learning visual representations that embraces multimodal data, driven by a combination of intra- and inter-modal similarity preservation objectives. Unlike existing visual pre-training methods, which solve a proxy prediction task in a single domain, our method exploits intrinsic data properties within each modality and semantic information from cross-modal correlation simultaneously, hence improving the quality of learned visual representations. By including multimodal training in a unified framework with different types of contrastive losses, our method can learn more powerful and generic visual features. We first train our model on COCO and evaluate the learned visual representations on various downstream tasks including image classification, object detection, and instance segmentation. For example, the visual representations pre-trained on COCO by our method achieve state-of-the-art top-1 validation accuracy of $55.3\%$ on ImageNet classification, under the common transfer protocol. We also evaluate our method on the large-scale Stock images dataset and show its effectiveness on multi-label image tagging, and cross-modal retrieval tasks.
Video classification and analysis is always a popular and challenging field in computer vision. It is more than just simple image classification due to the correlation with respect to the semantic contents of subsequent frames brings difficulties for video analysis. In this literature review, we summarized some state-of-the-art methods for multi-label video classification. Our goal is first to experimentally research the current widely used architectures, and then to develop a method to deal with the sequential data of frames and perform multi-label classification based on automatic content detection of video.
Video and image quality assessment has long been projected as a regression problem, which requires predicting a continuous quality score given an input stimulus. However, recent efforts have shown that accurate quality score regression on real-world user-generated content (UGC) is a very challenging task. To make the problem more tractable, we propose two new methods - binary, and ordinal classification - as alternatives to evaluate and compare no-reference quality models at coarser levels. Moreover, the proposed new tasks convey more practical meaning on perceptually optimized UGC transcoding, or for preprocessing on media processing platforms. We conduct a comprehensive benchmark experiment of popular no-reference quality models on recent in-the-wild picture and video quality datasets, providing reliable baselines for both evaluation methods to support further studies. We hope this work promotes coarse-grained perceptual modeling and its applications to efficient UGC processing.