Chest imaging plays an essential role in diagnosing and predicting patients with COVID-19 with evidence of worsening respiratory status. Many deep learning-based diagnostic models for pneumonia have been developed to enable computer-aided diagnosis. However, the long training and inference time make them inflexible. In addition, the lack of interpretability reduces their credibility in clinical medical practice. This paper presents CMT, a model with interpretability and rapid recognition of pneumonia, especially COVID-19 positive. Multiple convolutional layers in CMT are first used to extract features in CXR images, and then Transformer is applied to calculate the possibility of each symptom. To improve the model's generalization performance and to address the problem of sparse medical image data, we propose Feature Fusion Augmentation (FFA), a plug-and-play method for image augmentation. It fuses the features of the two images to varying degrees to produce a new image that does not deviate from the original distribution. Furthermore, to reduce the computational complexity and accelerate the convergence, we propose Multilevel Multi-Head Self-Attention (MMSA), which computes attention on different levels to establish the relationship between global and local features. It significantly improves the model performance while substantially reducing its training and inference time. Experimental results on the largest COVID-19 dataset show the proposed CMT has state-of-the-art performance. The effectiveness of FFA and MMSA is demonstrated in the ablation experiments. In addition, the weights and feature activation maps of the model inference process are visualized to show the CMT's interpretability.
Conditional image generation is an active research topic including text2image and image translation. Recently image manipulation with linguistic instruction brings new challenges of multimodal conditional generation. However, traditional conditional image generation models mainly focus on generating high-quality and visually realistic images, and lack resolving the partial consistency between image and instruction. To address this issue, we propose an Increment Reasoning Generative Adversarial Network (IR-GAN), which aims to reason the consistency between visual increment in images and semantic increment in instructions. First, we introduce the word-level and instruction-level instruction encoders to learn user's intention from history-correlated instructions as semantic increment. Second, we embed the representation of semantic increment into that of source image for generating target image, where source image plays the role of referring auxiliary. Finally, we propose a reasoning discriminator to measure the consistency between visual increment and semantic increment, which purifies user's intention and guarantees the good logic of generated target image. Extensive experiments and visualization conducted on two datasets show the effectiveness of IR-GAN.
Spatial transcriptomics (ST) is essential for understanding diseases and developing novel treatments. It measures gene expression of each fine-grained area (i.e., different windows) in the tissue slide with low throughput. This paper proposes an Exemplar Guided Network (EGN) to accurately and efficiently predict gene expression directly from each window of a tissue slide image. We apply exemplar learning to dynamically boost gene expression prediction from nearest/similar exemplars of a given tissue slide image window. Our EGN framework composes of three main components: 1) an extractor to structure a representation space for unsupervised exemplar retrievals; 2) a vision transformer (ViT) backbone to progressively extract representations of the input window; and 3) an Exemplar Bridging (EB) block to adaptively revise the intermediate ViT representations by using the nearest exemplars. Finally, we complete the gene expression prediction task with a simple attention-based prediction block. Experiments on standard benchmark datasets indicate the superiority of our approach when comparing with the past state-of-the-art (SOTA) methods.
This paper presents WALDO (WArping Layer-Decomposed Objects), a novel approach to the prediction of future video frames from past ones. Individual images are decomposed into multiple layers combining object masks and a small set of control points. The layer structure is shared across all frames in each video to build dense inter-frame connections. Complex scene motions are modeled by combining parametric geometric transformations associated with individual layers, and video synthesis is broken down into discovering the layers associated with past frames, predicting the corresponding transformations for upcoming ones and warping the associated object regions accordingly, and filling in the remaining image parts. Extensive experiments on the Cityscapes (resp. KITTI) dataset show that WALDO significantly outperforms prior works with, e.g., 3, 27, and 51% (resp. 5, 20 and 11%) relative improvement in SSIM, LPIPS and FVD metrics. Code, pretrained models, and video samples synthesized by our approach can be found in the project webpage https://16lemoing.github.io/waldo.
In recent years, the industrial sector has evolved towards its fourth revolution. The quality control domain is particularly interested in advanced machine learning for computer vision anomaly detection. Nevertheless, several challenges have to be faced, including imbalanced datasets, the image complexity, and the zero-false-negative (ZFN) constraint to guarantee the high-quality requirement. This paper illustrates a use case for an industrial partner, where Printed Circuit Board Assembly (PCBA) images are first reconstructed with a Vector Quantized Generative Adversarial Network (VQGAN) trained on normal products. Then, several multi-level metrics are extracted on a few normal and abnormal images, highlighting anomalies through reconstruction differences. Finally, a classifer is trained to build a composite anomaly score thanks to the metrics extracted. This three-step approach is performed on the public MVTec-AD datasets and on the partner PCBA dataset, where it achieves a regular accuracy of 95.69% and 87.93% under the ZFN constraint.
The photometric stereo (PS) problem consists in reconstructing the 3D-surface of an object, thanks to a set of photographs taken under different lighting directions. In this paper, we propose a multi-scale architecture for PS which, combined with a new dataset, yields state-of-the-art results. Our proposed architecture is flexible: it permits to consider a variable number of images as well as variable image size without loss of performance. In addition, we define a set of constraints to allow the generation of a relevant synthetic dataset to train convolutional neural networks for the PS problem. Our proposed dataset is much larger than pre-existing ones, and contains many objects with challenging materials having anisotropic reflectance (e.g. metals, glass). We show on publicly available benchmarks that the combination of both these contributions drastically improves the accuracy of the estimated normal field, in comparison with previous state-of-the-art methods.
Digital images are commonly represented as regular 2D arrays, so pixels are organized in form of a matrix addressed by integers. However, there are many image processing operations, such as rotation or motion compensation, that produce pixels at non-integer positions. Typically, image reconstruction techniques cannot handle samples at non-integer positions. In this paper, we propose to use triangulation-based reconstruction as initial estimate that is later refined by a novel adaptive denoising framework. Simulations reveal that improvements of up to more than 1.8 dB (in terms of PSNR) are achieved with respect to the initial estimate.
Image super-resolution (SR) techniques are used to generate a high-resolution image from a low-resolution image. Until now, deep generative models such as autoregressive models and Generative Adversarial Networks (GANs) have proven to be effective at modelling high-resolution images. Models based on Variational Autoencoders (VAEs) have often been criticized for their feeble generative performance, but with new advancements such as VDVAE (very deep VAE), there is now strong evidence that deep VAEs have the potential to outperform current state-of-the-art models for high-resolution image generation. In this paper, we introduce VDVAE-SR, a new model that aims to exploit the most recent deep VAE methodologies to improve upon image super-resolution using transfer learning on pretrained VDVAEs. Through qualitative and quantitative evaluations, we show that the proposed model is competitive with other state-of-the-art methods.
We propose a novel method to generate underwater object imagery that is acoustically compliant with that generated by side-scan sonar using the Unreal Engine. We describe the process to develop, tune, and generate imagery to provide representative images for use in training automated target recognition (ATR) and machine learning algorithms. The methods provide visual approximations for acoustic effects such as back-scatter noise and acoustic shadow, while allowing fast rendering with C++ actor in UE for maximizing the size of potential ATR training datasets. Additionally, we provide analysis of its utility as a replacement for actual sonar imagery or physics-based sonar data.
One major problem of objective Image Quality Assessment (IQA) methods is the lack of linearity of their quality estimates with respect to scores expressed by human subjects. For this reason, usually IQA metrics undergo a calibration process based on subjective quality examples. However, example-based training makes generalization problematic, hampering result comparison across different applications and operative conditions. In this paper, new Full Reference (FR) techniques, providing estimates linearly correlated with human scores without using calibration are introduced. To reach this objective, these techniques are deeply rooted on principles and theoretical constraints. Restricting the interest on the IQA of the set of natural images, it is first recognized that application of estimation theory and psycho physical principles to images degraded by Gaussian blur leads to a so-called canonical IQA method, whose estimates are not only highly linearly correlated to subjective scores, but are also straightforwardly related to the Viewing Distance (VD). Then, it is shown that mainstream IQA methods can be reconducted to the canonical method applying a preliminary metric conversion based on a unique specimen image. The application of this scheme is then extended to a significant class of degraded images other than Gaussian blur, including noisy and compressed images. The resulting calibration-free FR IQA methods are suited for applications where comparability and interoperability across different imaging systems and on different VDs is a major requirement. A comparison of their statistical performance with respect to some conventional calibration prone methods is finally provided.