Rapid damage assessment is one of the core tasks that response organizations perform at the onset of a disaster to understand the scale of damage to infrastructures such as roads, bridges, and buildings. This work analyzes the usefulness of social media imagery content to perform rapid damage assessment during a real-world disaster. An automatic image processing system, which was activated in collaboration with a volunteer response organization, processed ~280K images to understand the extent of damage caused by the disaster. The system achieved an accuracy of 76% computed based on the feedback received from the domain experts who analyzed ~29K system-processed images during the disaster. An extensive error analysis reveals several insights and challenges faced by the system, which are vital for the research community to advance this line of research.
Food image recognition is one of the promising applications of visual object recognition in computer vision. In this study, a small-scale dataset consisting of 5822 images of ten categories and a five-layer CNN was constructed to recognize these images. The bag-of-features (BoF) model coupled with support vector machine was first tested as comparison, resulting in an overall accuracy of 56%, while the CNN performed much better with an overall accuracy of 74%. Data expansion techniques were applied to increase the size of training images, which achieved a significantly improved accuracy of more than 90% and prevent the overfitting issue that occurred to the CNN without using data expansion. Further improvement is within reach by collecting more images and optimizing the network architecture and relevant hyper-parameters.
Focal-plane Sensor-processor (FPSP) is a next-generation camera technology which enables every pixel on the sensor chip to perform computation in parallel, on the focal plane where the light intensity is captured. SCAMP-5 is a general-purpose FPSP used in this work and it carries out computations in the analog domain before analog to digital conversion. By extracting features from the image on the focal plane, data which is digitized and transferred is reduced. As a consequence, SCAMP-5 offers a high frame rate while maintaining low energy consumption. Here, we present BIT-VO, which is, to the best of our knowledge, the first 6 Degrees of Freedom visual odometry algorithm which utilises the FPSP. Our entire system operates at 300 FPS in a natural scene, using binary edges and corner features detected by the SCAMP-5.
Medical imaging is an essential tool in many areas of medical applications, used for both diagnosis and treatment. However, reading medical images and making diagnosis or treatment recommendations require specially trained medical specialists. The current practice of reading medical images is labor-intensive, time-consuming, costly, and error-prone. It would be more desirable to have a computer-aided system that can automatically make diagnosis and treatment recommendations. Recent advances in deep learning enable us to rethink the ways of clinician diagnosis based on medical images. In this thesis, we will introduce 1) mammograms for detecting breast cancers, the most frequently diagnosed solid cancer for U.S. women, 2) lung CT images for detecting lung cancers, the most frequently diagnosed malignant cancer, and 3) head and neck CT images for automated delineation of organs at risk in radiotherapy. First, we will show how to employ the adversarial concept to generate the hard examples improving mammogram mass segmentation. Second, we will demonstrate how to use the weakly labeled data for the mammogram breast cancer diagnosis by efficiently design deep learning for multi-instance learning. Third, the thesis will walk through DeepLung system which combines deep 3D ConvNets and GBM for automated lung nodule detection and classification. Fourth, we will show how to use weakly labeled data to improve existing lung nodule detection system by integrating deep learning with a probabilistic graphic model. Lastly, we will demonstrate the AnatomyNet which is thousands of times faster and more accurate than previous methods on automated anatomy segmentation.
The 3D documentation of the tunnel surface during construction requires fast and robust measurement systems. In the solution proposed in this paper, during tunnel advance a single camera is taking pictures of the tunnel surface from several positions. The recorded images are automatically processed to gain a 3D tunnel surface model. Image acquisition is realized by the tunneling/advance/driving personnel close to the tunnel face (= the front end of the advance). Based on the following fully automatic analysis/evaluation, a decision on the quality of the outbreak can be made within a few minutes. This paper describes the image recording system and conditions as well as the stereo-photogrammetry based workflow for the continuously merged dense 3D reconstruction of the entire advance region. Geo-reference is realized by means of signalized targets that are automatically detected in the images. We report on the results of recent testing under real construction conditions, and conclude with prospects for further development in terms of on-site performance.
Real-world settings often do not allow acquisition of high-resolution volumetric images for accurate morphological assessment and diagnostic. In clinical practice it is frequently common to acquire only sparse data (e.g. individual slices) for initial diagnostic decision making. Thereby, physicians rely on their prior knowledge (or mental maps) of the human anatomy to extrapolate the underlying 3D information. Accurate mental maps require years of anatomy training, which in the first instance relies on normative learning, i.e. excluding pathology. In this paper, we leverage Bayesian Deep Learning and environment mapping to generate full volumetric anatomy representations from none to a small, sparse set of slices. We evaluate proof of concept implementations based on Generative Query Networks (GQN) and Conditional BRUNO using abdominal CT and brain MRI as well as in a clinical application involving sparse, motion-corrupted MR acquisition for fetal imaging. Our approach allows to reconstruct 3D volumes from 1 to 4 tomographic slices, with a SSIM of 0.7+ and cross-correlation of 0.8+ compared to the 3D ground truth.
The current industry practice for 24-hour outdoor imaging is to use a silicon camera supplemented with near-infrared (NIR) illumination. This will result in color images with poor contrast at daytime and absence of chrominance at nighttime. For this dilemma, all existing solutions try to capture RGB and NIR images separately. However, they need additional hardware support and suffer from various drawbacks, including short service life, high price, specific usage scenario, etc. In this paper, we propose a novel and integrated enhancement solution that produces clear color images, whether at abundant sunlight daytime or extremely low-light nighttime. Our key idea is to separate the VIS and NIR information from mixed signals, and enhance the VIS signal adaptively with the NIR signal as assistance. To this end, we build an optical system to collect a new VIS-NIR-MIX dataset and present a physically meaningful image processing algorithm based on CNN. Extensive experiments show outstanding results, which demonstrate the effectiveness of our solution.
Generative Adversarial Networks (GANs) have made great progress in synthesizing realistic images in recent years. However, they are often trained on image datasets with either too few samples or too many classes belonging to different data distributions. Consequently, GANs are prone to underfitting or overfitting, making the analysis of them difficult and constrained. Therefore, in order to conduct a thorough study on GANs while obviating unnecessary interferences introduced by the datasets, we train them on artificial datasets where there are infinitely many samples and the real data distributions are simple, high-dimensional and have structured manifolds. Moreover, the generators are designed such that optimal sets of parameters exist. Empirically, we find that under various distance measures, the generator fails to learn such parameters with the GAN training procedure. We also find that training mixtures of GANs leads to more performance gain compared to increasing the network depth or width when the model complexity is high enough. Our experimental results demonstrate that a mixture of generators can discover different modes or different classes automatically in an unsupervised setting, which we attribute to the distribution of the generation and discrimination tasks across multiple generators and discriminators. As an example of the generalizability of our conclusions to realistic datasets, we train a mixture of GANs on the CIFAR-10 dataset and our method significantly outperforms the state-of-the-art in terms of popular metrics, i.e., Inception Score (IS) and Fr\'echet Inception Distance (FID).
Additive manufacturing has the potential to become a standard method for manufacturing products, and product information is indispensable for the item distribution system. While most products are given barcodes to the exterior surfaces, research on embedding barcodes inside products is underway. This is because additive manufacturing makes it possible to carry out manufacturing and information adding at the same time, and embedding information inside does not impair the exterior appearance of the product. However, products that have not been embedded information can not be identified, and embedded information can not be rewritten later. In this study, we have developed a product identification system that does not require embedding barcodes inside. This system uses a transmission image of the product which contains information of each product such as different inner support structures and manufacturing errors. We have shown through experiments that if datasets of transmission images are available, objects can be identified with an accuracy of over 90%. This result suggests that our approach can be useful for identifying objects without embedded information.
Convolutional Neural Networks demonstrate high performance on ImageNet Large-Scale Visual Recognition Challenges contest. Nevertheless, the published results only show the overall performance for all image classes. There is no further analysis why certain images get worse results and how they could be improved. In this paper, we provide deep performance analysis based on different types of images and point out the weaknesses of convolutional neural networks through experiment. We design a novel multiple paths convolutional neural network, which feeds different versions of images into separated paths to learn more comprehensive features. This model has better presentation for image than the traditional single path model. We acquire better classification results on complex validation set on both top 1 and top 5 scores than the best ILSVRC 2013 classification model.