We propose a mechanism to reconstruct part annotated 3D point clouds of objects given just a single input image. We demonstrate that jointly training for both reconstruction and segmentation leads to improved performance in both the tasks, when compared to training for each task individually. The key idea is to propagate information from each task so as to aid the other during the training procedure. Towards this end, we introduce a location-aware segmentation loss in the training regime. We empirically show the effectiveness of the proposed loss in generating more faithful part reconstructions while also improving segmentation accuracy. We thoroughly evaluate the proposed approach on different object categories from the ShapeNet dataset to obtain improved results in reconstruction as well as segmentation.
Arbitrary-shaped text detection is an important and challenging task in computer vision. Most existing methods require heavy data labeling efforts to produce polygon-level text region labels for supervised training. In order to reduce the cost in data labeling, we study weakly-supervised arbitrary-shaped text detection for combining various weak supervision forms (e.g., image-level tags, coarse, loose and tight bounding boxes), which are far easier for annotation. We propose an Expectation-Maximization (EM) based weakly-supervised learning framework to train an accurate arbitrary-shaped text detector using only a small amount of polygon-level annotated data combined with a large amount of weakly annotated data. Meanwhile, we propose a contour-based arbitrary-shaped text detector, which is suitable for incorporating weakly-supervised learning. Extensive experiments on three arbitrary-shaped text benchmarks (CTW1500, Total-Text and ICDAR-ArT) show that (1) using only 10% strongly annotated data and 90% weakly annotated data, our method yields comparable performance to state-of-the-art methods, (2) with 100% strongly annotated data, our method outperforms existing methods on all three benchmarks. We will make the weakly annotated datasets publicly available in the future.
Transformer models have recently achieved impressive performance on NLP tasks, owing to new algorithms for self-supervised pre-training on very large text corpora. In contrast, recent literature suggests that simple average word models outperform more complicated language models, e.g., RNNs and Transformers, on cross-modal image/text search tasks on standard benchmarks, like MS COCO. In this paper, we show that dataset scale and training strategy are critical and demonstrate that transformer-based cross-modal embeddings outperform word average and RNN-based embeddings by a large margin, when trained on a large dataset of e-commerce product image-title pairs.
Plenoptic cameras offer a cost effective solution to capture light fields by multiplexing multiple views on a single image sensor. However, the high angular resolution is achieved at the expense of reducing the spatial resolution of each view by orders of magnitude compared to the raw sensor image. While light field super-resolution is still at an early stage, the field of single image super-resolution (SISR) has recently known significant advances with the use of deep learning techniques. This paper describes a simple framework allowing us to leverage state-of-the-art SISR techniques into light fields, while taking into account specific light field geometrical constraints. The idea is to first compute a representation compacting most of the light field energy into as few components as possible. This is achieved by aligning the light field using optical flows and then by decomposing the aligned light field using singular value decomposition (SVD). The principal basis captures the information that is coherent across all the views, while the other basis contain the high angular frequencies. Super-resolving this principal basis using an SISR method allows us to super-resolve all the information that is coherent across the entire light field. This framework allows the proposed light field super-resolution method to inherit the benefits of the SISR method used. Experimental results show that the proposed method is competitive, and most of the time superior, to recent light field super-resolution methods in terms of both PSNR and SSIM quality metrics, with a lower complexity.
We present a randomized controlled trial for a model-in-the-loop regression task, with the goal of measuring the extent to which (1) good explanations of model predictions increase human accuracy, and (2) faulty explanations decrease human trust in the model. We study explanations based on visual saliency in an image-based age prediction task for which humans and learned models are individually capable but not highly proficient and frequently disagree. Our experimental design separates model quality from explanation quality, and makes it possible to compare treatments involving a variety of explanations of varying levels of quality. We find that presenting model predictions improves human accuracy. However, visual explanations of various kinds fail to significantly alter human accuracy or trust in the model - regardless of whether explanations characterize an accurate model, an inaccurate one, or are generated randomly and independently of the input image. These findings suggest the need for greater evaluation of explanations in downstream decision making tasks, better design-based tools for presenting explanations to users, and better approaches for generating explanations.
Training robots to navigate diverse environments is a challenging problem as it involves the confluence of several different perception tasks such as mapping and localization, followed by optimal path-planning and control. Recently released photo-realistic simulators such as Habitat allow for the training of networks that output control actions directly from perception: agents use Deep Reinforcement Learning (DRL) to regress directly from the camera image to a control output in an end-to-end fashion. This is data-inefficient and can take several days to train on a GPU. Our paper tries to overcome this problem by separating the training of the perception and control neural nets and increasing the path complexity gradually using a curriculum approach. Specifically, a pre-trained twin Variational AutoEncoder (VAE) is used to compress RGBD (RGB & depth) sensing from an environment into a latent embedding, which is then used to train a DRL-based control policy. A*, a traditional path-planner is used as a guide for the policy and the distance between start and target locations is incrementally increased along the A* route, as training progresses. We demonstrate the efficacy of the proposed approach, both in terms of increased performance and decreased training times for the PointNav task in the Habitat simulation environment. This strategy of improving the training of direct-perception based DRL navigation policies is expected to hasten the deployment of robots of particular interest to industry such as co-bots on the factory floor and last-mile delivery robots.
We consider the reconstruction problem of video snapshot compressive imaging (SCI), which captures high-speed videos using a low-speed 2D sensor (detector). The underlying principle of SCI is to modulate sequential high-speed frames with different masks and then these encoded frames are integrated into a snapshot on the sensor and thus the sensor can be of low-speed. On one hand, video SCI enjoys the advantages of low-bandwidth, low-power and low-cost. On the other hand, applying SCI to large-scale problems (HD or UHD videos) in our daily life is still challenging and one of the bottlenecks lies in the reconstruction algorithm. Exiting algorithms are either too slow (iterative optimization algorithms) or not flexible to the encoding process (deep learning based end-to-end networks). In this paper, we develop fast and flexible algorithms for SCI based on the plug-and-play (PnP) framework. In addition to the PnP-ADMM method, we further propose the PnP-GAP (generalized alternating projection) algorithm with a lower computational workload. We first employ the image deep denoising priors to show that PnP can recover a UHD color video with 30 frames from a snapshot measurement. Since videos have strong temporal correlation, by employing the video deep denoising priors, we achieve a significant improvement in the results. Furthermore, we extend the proposed PnP algorithms to the color SCI system using mosaic sensors, where each pixel only captures the red, green or blue channels. A joint reconstruction and demosaicing paradigm is developed for flexible and high quality reconstruction of color video SCI systems. Extensive results on both simulation and real datasets verify the superiority of our proposed algorithm.
While the field of deep tomographic reconstruction has been advancing rapidly since 2016, there are constant debates and major challenges with the recently published PNAS paper on instabilities of deep learning in image reconstruction as a primary example, in which three kinds of unstable phenomena are demonstrated: (1) tiny perturbation on input generating strong output artifacts, (2) small structural features going undetected, and (3) increased input data leading to decreased performance. In this article, we show that key algorithmic ingredients of analytic inversion, compressed sensing, iterative reconstruction, and deep learning can be synergized to stabilize deep neural networks for optimal tomographic image reconstruction. With the same or similar datasets used in the PNAS paper and relative to the same state of the art compressed sensing algorithm, our proposed analytic, compressed, iterative deep (ACID) network produces superior imaging performance that are both accurate and robust with respect to noise, under adversarial attack, and as the number of input data is increased. We believe that deep tomographic reconstruction networks can be designed to produce accurate and robust results, improve clinical and other important applications, and eventually dominate the tomographic imaging field.
Chest X-ray is the first imaging technique that plays an important role in the diagnosis of COVID-19 disease. Due to the high availability of large-scale annotated image datasets, great success has been achieved using convolutional neural networks (CNNs) for image recognition and classification. However, due to the limited availability of annotated medical images, the classification of medical images remains the biggest challenge in medical diagnosis. Thanks to transfer learning, an effective mechanism that can provide a promising solution by transferring knowledge from generic object recognition tasks to domain-specific tasks. In this paper, we validate and adapt our previously developed CNN, called Decompose, Transfer, and Compose (DeTraC), for the classification of COVID-19 chest X-ray images. DeTraC can deal with any irregularities in the image dataset by investigating its class boundaries using a class decomposition mechanism. The experimental results showed the capability of DeTraC in the detection of COVID-19 cases from a comprehensive image dataset collected from several hospitals around the world. High accuracy of 95.12% (with a sensitivity of 97.91%, a specificity of 91.87%, and a precision of 93.36%) was achieved by DeTraC in the detection of COVID-19 X-ray images from normal, and severe acute respiratory syndrome cases.
Recognition of document images have important applications in restoring old and classical texts. The problem involves quality improvement before passing it to a properly trained OCR to get accurate recognition of the text. The image enhancement and quality improvement constitute important steps as subsequent recognition depends upon the quality of the input image. There are scenarios when high resolution images are not available and our experiments show that the OCR accuracy reduces significantly with decrease in the spatial resolution of document images. Thus the only option is to improve the resolution of such document images. The goal is to construct a high resolution image, given a single low resolution binary image, which constitutes the problem of single image super-resolution. Most of the previous work in super-resolution deal with natural images which have more information-content than the document images. Here, we use Convolution Neural Network to learn the mapping between low and the corresponding high resolution images. We experiment with different number of layers, parameter settings and non-linear functions to build a fast end-to-end framework for document image super-resolution. Our proposed model shows a very good PSNR improvement of about 4 dB on 75 dpi Tamil images, resulting in a 3 % improvement of word level accuracy by the OCR. It takes less time than the recent sparse based natural image super-resolution technique, making it useful for real-time document recognition applications.