1.VL-BERT: Pre-training of Generic Visual-Linguistic Representations ⬇️
We introduce a new pre-trainable generic representation for visual-linguistic tasks, called Visual-Linguistic BERT (VL-BERT for short). VL-BERT adopts the simple yet powerful Transformer model as the backbone, and extends it to take both visual and linguistic embedded features as input. In it, each element of the input is either of a word from the input sentence, or a region-of-interest (RoI) from the input image. It is designed to fit for most of the vision-and-language downstream tasks. To better exploit the generic representation, we pre-train VL-BERT on massive-scale Conceptual Captions dataset with three tasks: masked language modeling with visual clues, masked RoI classification with linguistic clues, and sentence-image relationship prediction. Extensive empirical analysis demonstrates that the pre-training procedure can better align the visual-linguistic clues and benefit the downstream tasks, such as visual question answering, visual commonsense reasoning and referring expression comprehension. It is worth noting that VL-BERT achieved the first place of single model on the leaderboard of the VCR benchmark.
2.Sequential Latent Spaces for Modeling the Intention During Diverse Image Captioning ⬇️
Diverse and accurate vision+language modeling is an important goal to retain creative freedom and maintain user engagement. However, adequately capturing the intricacies of diversity in language models is challenging. Recent works commonly resort to latent variable models augmented with more or less supervision from object detectors or part-of-speech tags. Common to all those methods is the fact that the latent variable either only initializes the sentence generation process or is identical across the steps of generation. Both methods offer no fine-grained control. To address this concern, we propose Seq-CVAE which learns a latent space for every word position. We encourage this temporal latent space to capture the 'intention' about how to complete the sentence by mimicking a representation which summarizes the future. We illustrate the efficacy of the proposed approach to anticipate the sentence continuation on the challenging MSCOCO dataset, significantly improving diversity metrics compared to baselines while performing on par w.r.t sentence quality.
3.ViCo: Word Embeddings from Visual Co-occurrences ⬇️
We propose to learn word embeddings from visual co-occurrences. Two words co-occur visually if both words apply to the same image or image region. Specifically, we extract four types of visual co-occurrences between object and attribute words from large-scale, textually-annotated visual databases like VisualGenome and ImageNet. We then train a multi-task log-bilinear model that compactly encodes word "meanings" represented by each co-occurrence type into a single visual word-vector. Through unsupervised clustering, supervised partitioning, and a zero-shot-like generalization analysis we show that our word embeddings complement text-only embeddings like GloVe by better representing similarities and differences between visual concepts that are difficult to obtain from text corpora alone. We further evaluate our embeddings on five downstream applications, four of which are vision-language tasks. Augmenting GloVe with our embeddings yields gains on all tasks. We also find that random embeddings perform comparably to learned embeddings on all supervised vision-language tasks, contrary to conventional wisdom.
4.Compositional Video Prediction ⬇️
We present an approach for pixel-level future prediction given an input image of a scene. We observe that a scene is comprised of distinct entities that undergo motion and present an approach that operationalizes this insight. We implicitly predict future states of independent entities while reasoning about their interactions, and compose future video frames using these predicted states. We overcome the inherent multi-modality of the task using a global trajectory-level latent random variable, and show that this allows us to sample diverse and plausible futures. We empirically validate our approach against alternate representations and ways of incorporating multi-modality. We examine two datasets, one comprising of stacked objects that may fall, and the other containing videos of humans performing activities in a gym, and show that our approach allows realistic stochastic video prediction across these diverse settings. See this https URL for video predictions.
5.Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement ⬇️
Generic Image recognition is a fundamental and fairly important visual problem in computer vision. One of the major challenges of this task lies in the fact that single image usually has multiple objects inside while the labels are still one-hot, another one is noisy and sometimes missing labels when annotated by humans. In this paper, we focus on tackling these challenges accompanying with two different image recognition problems: multi-model ensemble and noisy data recognition with a unified framework. As is well-known, usually the best performing deep neural models are ensembles of multiple base-level networks, as it can mitigate the variation or noise containing in the dataset. Unfortunately, the space required to store these many networks, and the time required to execute them at runtime, prohibit their use in applications where test sets are large (e.g., ImageNet). In this paper, we present a method for compressing large, complex trained ensembles into a single network, where the knowledge from a variety of trained deep neural networks (DNNs) is distilled and transferred to a single DNN. In order to distill diverse knowledge from different trained (teacher) models, we propose to use adversarial-based learning strategy where we define a block-wise training loss to guide and optimize the predefined student network to recover the knowledge in teacher models, and to promote the discriminator network to distinguish teacher vs. student features simultaneously. Extensive experiments on CIFAR-10/100, SVHN, ImageNet and iMaterialist Challenge Dataset demonstrate the effectiveness of our MEAL method. On ImageNet, our ResNet-50 based MEAL achieves top-1/5 21.79%/5.99% val error, which outperforms the original model by 2.06%/1.14%. On iMaterialist Challenge Dataset, our MEAL obtains a remarkable improvement of top-3 1.15% (official evaluation metric) on a strong baseline model of ResNet-101.
6.Predicting Animation Skeletons for 3D Articulated Models via Volumetric Nets ⬇️
We present a learning method for predicting animation skeletons for input 3D models of articulated characters. In contrast to previous approaches that fit pre-defined skeleton templates or predict fixed sets of joints, our method produces an animation skeleton tailored for the structure and geometry of the input 3D model. Our architecture is based on a stack of hourglass modules trained on a large dataset of 3D rigged characters mined from the web. It operates on the volumetric representation of the input 3D shapes augmented with geometric shape features that provide additional cues for joint and bone locations. Our method also enables intuitive user control of the level-of-detail for the output skeleton. Our evaluation demonstrates that our approach predicts animation skeletons that are much more similar to the ones created by humans compared to several alternatives and baselines.
7.EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition ⬇️
We focus on multi-modal fusion for egocentric action recognition, and propose a novel architecture for multi-modal temporal-binding, i.e. the combination of modalities within a range of temporal offsets. We train the architecture with three modalities -- RGB, Flow and Audio -- and combine them with mid-level fusion alongside sparse temporal sampling of fused representations. In contrast with previous works, modalities are fused before temporal aggregation, with shared modality and fusion weights over time. Our proposed architecture is trained end-to-end, outperforming individual modalities as well as late-fusion of modalities.
We demonstrate the importance of audio in egocentric vision, on per-class basis, for identifying actions as well as interacting objects. Our method achieves state of the art results on both the seen and unseen test sets of the largest egocentric dataset: EPIC-Kitchens, on all metrics using the public leaderboard.
8.Noise Flow: Noise Modeling with Conditional Normalizing Flows ⬇️
Modeling and synthesizing image noise is an important aspect in many computer vision applications. The long-standing additive white Gaussian and heteroscedastic (signal-dependent) noise models widely used in the literature provide only a coarse approximation of real sensor noise. This paper introduces Noise Flow, a powerful and accurate noise model based on recent normalizing flow architectures. Noise Flow combines well-established basic parametric noise models (e.g., signal-dependent noise) with the flexibility and expressiveness of normalizing flow networks. The result is a single, comprehensive, compact noise model containing fewer than 2500 parameters yet able to represent multiple cameras and gain factors. Noise Flow dramatically outperforms existing noise models, with 0.42 nats/pixel improvement over the camera-calibrated noise level functions, which translates to 52% improvement in the likelihood of sampled noise. Noise Flow represents the first serious attempt to go beyond simple parametric models to one that leverages the power of deep learning and data-driven noise distributions.
9.Scoot: A Perceptual Metric for Facial Sketches ⬇️
While it is trivial for humans to quickly assess the perceptual similarity between two images, the underlying mechanism is thought to be quite complex. Despite this, the most widely adopted perceptual metrics today, such as SSIM and FSIM, are simple, shallow functions, and fail to consider many factors of human perception. Recently, the facial modelling community has observed that the inclusion of both structure and texture has a significant positive benefit for face sketch synthesis (FSS). But how perceptual are these so-called "perceptual features"? Which elements are critical for their success? In this paper, we design a perceptual metric, called Structure Co-Occurrence Texture (Scoot), which simultaneously considers the blocklevel spatial structure and co-occurrence texture statistics. To test the quality of metrics, we propose three novel metameasures based on various reliable properties. Extensive experiments verify that our Scoot metric exceeds the performance of prior work. Besides, we built the first large scale (152k judgments) human-perception-based sketch database that can evaluate how well a metric is consistent with human perception. Our results suggest that "spatial structure" and "co-occurrence texture" are two generally applicable perceptual features in face sketch synthesis.
10.Saliency Methods for Explaining Adversarial Attacks ⬇️
In this work, we aim to explain the classifications of adversary images using saliency methods. Saliency methods explain individual classification decisions of neural networks by creating saliency maps. All saliency methods were proposed for explaining correct predictions. Recent research shows that many proposed saliency methods fail to explain the predictions. Notably, the Guided Backpropagation (GuidedBP) is essentially doing (partial) image recovery. In our work, our numerical analysis shows the saliency maps created by GuidedBP do contain class-discriminative information. We propose a simple and efficient way to enhance the created saliency maps. The proposed enhanced GuidedBP is the state-of-the-art saliency method to explain adversary classifications.
11.Indoor Depth Completion with Boundary Consistency and Self-Attention ⬇️
Depth estimation features are helpful for 3D recognition. Commodity-grade depth cameras are able to capture depth and color image in real-time. However, glossy, transparent or distant surface cannot be scanned properly by the sensor. As a result, enhancement and restoration from sensing depth is an important task. Depth completion aims at filling the holes that sensors fail to detect, which is still a complex task for machine to learn. Traditional hand-tuned methods have reached their limits, while neural network based methods tend to copy and interpolate the output from surrounding depth values. This leads to blurred boundaries, and structures of the depth map are lost. Consequently, our main work is to design an end-to-end network improving completion depth maps while maintaining edge clarity. We utilize self-attention mechanism, previously used in image inpainting fields, to extract more useful information in each layer of convolution so that the complete depth map is enhanced. In addition, we propose boundary consistency concept to enhance the depth map quality and structure. Experimental results validate the effectiveness of our self-attention and boundary consistency schema, which outperforms previous state-of-the-art depth completion work on Matterport3D dataset.
12.Deep Green Function Convolution for Improving Saliency in Convolutional Neural Networks ⬇️
Current saliency methods require to learn large scale regional features using small convolutional kernels, which is not possible with a simple feed-forward network. Some methods solve this problem by using segmentation into superpixels while others downscale the image through the network and rescale it back to its original size. The objective of this paper is to show that saliency convolutional neural networks (CNN) can be improved by using a Green's function convolution (GFC) to extrapolate edges features into salient regions. The GFC acts as a gradient integrator, allowing to produce saliency features from thin edge-like features directly inside the CNN. Hence, we propose the gradient integration and sum (GIS) layer that combines the edges features with the saliency features. Using the HED and DSS architecture, we demonstrated that adding a GIS layer near the network's output allows to reduce the sensitivity to the parameter initialization and overfitting, thus improving the repeatability of the training. By adding a GIS layer to the state-of-the-art DSS model, there is an increase of 1.6% for the F-measure on the DUT-OMRON dataset, with only 10ms of additional computation time. The GIS layer further allows the network to perform significantly better in the case of highly noisy images or low-brightness images. In fact, we observed an F-measure improvement of 5.2% on noisy images and 2.8% on low-light images. Since the GIS layer is model agnostic, it can be implemented inside different fully convolutional networks, and it outperforms the denseCRF post-processing method and is 40 times faster. A major contribution of the current work is the first implementation of Green's function convolution inside a neural network, which allows the network, via very minor architectural changes and no additional parameters, to operate in the feature domain and in the gradient domain at the same time.
13.EGNet:Edge Guidance Network for Salient Object Detection ⬇️
Fully convolutional neural networks (FCNs) have shown their advantages in the salient object detection task. However, most existing FCNs-based methods still suffer from coarse object boundaries. In this paper, to solve this problem, we focus on the complementarity between salient edge information and salient object information. Accordingly, we present an edge guidance network (EGNet) for salient object detection with three steps to simultaneously model these two kinds of complementary information in a single network. In the first step, we extract the salient object features by a progressive fusion way. In the second step, we integrate the local edge information and global location information to obtain the salient edge features. Finally, to sufficiently leverage these complementary features, we couple the same salient edge features with salient object features at various resolutions. Benefiting from the rich edge information and location information in salient edge features, the fused features can help locate salient objects, especially their boundaries more accurately. Experimental results demonstrate that the proposed method performs favorably against the state-of-the-art methods on six widely used datasets without any pre-processing and post-processing. The source code is available at http: //mmcheng.net/egnet/.
14.Trajectory Space Factorization for Deep Video-Based 3D Human Pose Estimation ⬇️
Existing deep learning approaches on 3d human pose estimation for videos are either based on Recurrent or Convolutional Neural Networks (RNNs or CNNs). However, RNN-based frameworks can only tackle sequences with limited frames because sequential models are sensitive to bad frames and tend to drift over long sequences. Although existing CNN-based temporal frameworks attempt to address the sensitivity and drift problems by concurrently processing all input frames in the sequence, the existing state-of-the-art CNN-based framework is limited to 3d pose estimation of a single frame from a sequential input. In this paper, we propose a deep learning-based framework that utilizes matrix factorization for sequential 3d human poses estimation. Our approach processes all input frames concurrently to avoid the sensitivity and drift problems, and yet outputs the 3d pose estimates for every frame in the input sequence. More specifically, the 3d poses in all frames are represented as a motion matrix factorized into a trajectory bases matrix and a trajectory coefficient matrix. The trajectory bases matrix is precomputed from matrix factorization approaches such as Singular Value Decomposition (SVD) or Discrete Cosine Transform (DCT), and the problem of sequential 3d pose estimation is reduced to training a deep network to regress the trajectory coefficient matrix. We demonstrate the effectiveness of our framework on long sequences by achieving state-of-the-art performances on multiple benchmark datasets. Our source code is available at: this https URL.
15.Object detection on aerial imagery using CenterNet ⬇️
Detection and classification of objects in aerial imagery have several applications like urban planning, crop surveillance, and traffic surveillance. However, due to the lower resolution of the objects and the effect of noise in aerial images, extracting distinguishing features for the objects is a challenge. We evaluate CenterNet, a state of the art method for real-time 2D object detection, on the VisDrone2019 dataset. We evaluate the performance of the model with different backbone networks in conjunction with varying resolutions during training and testing.
16.Uncertainty-Guided Domain Alignment for Layer Segmentation in OCT Images ⬇️
Automatic and accurate segmentation for retinal and choroidal layers of Optical Coherence Tomography (OCT) is crucial for detection of various ocular diseases. However, because of the variations in different equipments, OCT data obtained from different manufacturers might encounter appearance discrepancy, which could lead to performance fluctuation to a deep neural network. In this paper, we propose an uncertainty-guided domain alignment method to aim at alleviating this problem to transfer discriminative knowledge across distinct domains. We disign a novel uncertainty-guided cross-entropy loss for boosting the performance over areas with high uncertainty. An uncertainty-guided curriculum transfer strategy is developed for the self-training (ST), which regards uncertainty as efficient and effective guidance to optimize the learning process in target domain. Adversarial learning with feature recalibration module (FRM) is applied to transfer informative knowledge from the domain feature spaces adaptively. The experiments on two OCT datasets show that the proposed methods can obtain significant segmentation improvements compared with the baseline models.
17.Progressive Face Super-Resolution via Attention to Facial Landmark ⬇️
Face Super-Resolution (SR) is a subfield of the SR domain that specifically targets the reconstruction of face images. The main challenge of face SR is to restore essential facial features without distortion. We propose a novel face SR method that generates photo-realistic 8x super-resolved face images with fully retained facial details. To that end, we adopt a progressive training method, which allows stable training by splitting the network into successive steps, each producing output with a progressively higher resolution. We also propose a novel facial attention loss and apply it at each step to focus on restoring facial attributes in greater details by multiplying the pixel difference and heatmap values. Lastly, we propose a compressed version of the state-of-the-art face alignment network (FAN) for landmark heatmap extraction. With the proposed FAN, we can extract the heatmaps suitable for face SR and also reduce the overall training time. Experimental results verify that our method outperforms state-of-the-art methods in both qualitative and quantitative measurements, especially in perceptual quality.
18.3C-Net: Category Count and Center Loss for Weakly-Supervised Action Localization ⬇️
Temporal action localization is a challenging computer vision problem with numerous real-world applications. Most existing methods require laborious frame-level supervision to train action localization models. In this work, we propose a framework, called 3C-Net, which only requires video-level supervision (weak supervision) in the form of action category labels and the corresponding count. We introduce a novel formulation to learn discriminative action features with enhanced localization capabilities. Our joint formulation has three terms: a classification term to ensure the separability of learned action features, an adapted multi-label center loss term to enhance the action feature discriminability and a counting loss term to delineate adjacent action sequences, leading to improved localization. Comprehensive experiments are performed on two challenging benchmarks: THUMOS14 and ActivityNet 1.2. Our approach sets a new state-of-the-art for weakly-supervised temporal action localization on both datasets. On the THUMOS14 dataset, the proposed method achieves an absolute gain of 4.6% in terms of mean average precision (mAP), compared to the state-of-the-art. Source code is available at this https URL.
19.Mask TextSpotter: An End-to-End Trainable Neural Network for Spotting Text with Arbitrary Shapes ⬇️
Unifying text detection and text recognition in an end-to-end training fashion has become a new trend for reading text in the wild, as these two tasks are highly relevant and complementary. In this paper, we investigate the problem of scene text spotting, which aims at simultaneous text detection and recognition in natural images. An end-to-end trainable neural network named as Mask TextSpotter is presented. Different from the previous text spotters that follow the pipeline consisting of a proposal generation network and a sequence-to-sequence recognition network, Mask TextSpotter enjoys a simple and smooth end-to-end learning procedure, in which both detection and recognition can be achieved directly from two-dimensional space via semantic segmentation. Further, a spatial attention module is proposed to enhance the performance and universality. Benefiting from the proposed two-dimensional representation on both detection and recognition, it easily handles text instances of irregular shapes, for instance, curved text. We evaluate it on four English datasets and one multi-language dataset, achieving consistently superior performance over state-of-the-art methods in both detection and end-to-end text recognition tasks. Moreover, we further investigate the recognition module of our method separately, which significantly outperforms state-of-the-art methods on both regular and irregular text datasets for scene text recognition.
20.Pro-Cam SSfM: Projector-Camera System for Structure and Spectral Reflectance from Motion ⬇️
In this paper, we propose a novel projector-camera system for practical and low-cost acquisition of a dense object 3D model with the spectral reflectance property. In our system, we use a standard RGB camera and leverage an off-the-shelf projector as active illumination for both the 3D reconstruction and the spectral reflectance estimation. We first reconstruct the 3D points while estimating the poses of the camera and the projector, which are alternately moved around the object, by combining multi-view structured light and structure-from-motion (SfM) techniques. We then exploit the projector for multispectral imaging and estimate the spectral reflectance of each 3D point based on a novel spectral reflectance estimation model considering the geometric relationship between the reconstructed 3D points and the estimated projector positions. Experimental results on several real objects demonstrate that our system can precisely acquire a dense 3D model with the full spectral reflectance property using off-the-shelf devices.
21.Multi-Stream Single Shot Spatial-Temporal Action Detection ⬇️
We present a 3D Convolutional Neural Networks (CNNs) based single shot detector for spatial-temporal action detection tasks. Our model includes: (1) two short-term appearance and motion streams, with single RGB and optical flow image input separately, in order to capture the spatial and temporal information for the current frame; (2) two long-term 3D ConvNet based stream, working on sequences of continuous RGB and optical flow images to capture the context from past frames. Our model achieves strong performance for action detection in video and can be easily integrated into any current two-stream action detection methods. We report a frame-mAP of 71.30% on the challenging UCF101-24 actions dataset, achieving the state-of-the-art result of the one-stage methods. To the best of our knowledge, our work is the first system that combined 3D CNN and SSD in action detection tasks.
22.Globally optimal registration of noisy point clouds ⬇️
Registration of 3D point clouds is a fundamental task in several applications of robotics and computer vision. While registration methods such as iterative closest point and variants are very popular, they are only locally optimal. There has been some recent work on globally optimal registration, but they perform poorly in the presence of noise in the measurements. In this work we develop a mixed integer programming-based approach for globally optimal registration that explicitly considers uncertainty in its optimization, and hence produces more accurate estimates. Furthermore, from a practical implementation perspective we develop a multi-step optimization that combines fast local methods with our accurate global formulation. Through extensive simulation and real world experiments we demonstrate improved performance over state-of-the-art methods for various level of noise and outliers in the data as well as for partial geometric overlap.
23.Multiple instance dense connected convolution neural network for aerial image scene classification ⬇️
With the development of deep learning, many state-of-the-art natural image scene classification methods have demonstrated impressive performance. While the current convolution neural network tends to extract global features and global semantic information in a scene, the geo-spatial objects can be located at anywhere in an aerial image scene and their spatial arrangement tends to be more complicated. One possible solution is to preserve more local semantic information and enhance feature propagation. In this paper, an end to end multiple instance dense connected convolution neural network (MIDCCNN) is proposed for aerial image scene classification. First, a 23 layer dense connected convolution neural network (DCCNN) is built and served as a backbone to extract convolution features. It is capable of preserving middle and low level convolution features. Then, an attention based multiple instance pooling is proposed to highlight the local semantics in an aerial image scene. Finally, we minimize the loss between the bag-level predictions and the ground truth labels so that the whole framework can be trained directly. Experiments on three aerial image datasets demonstrate that our proposed methods can outperform current baselines by a large margin.
24.U-Net Training with Instance-Layer Normalization ⬇️
Normalization layers are essential in a Deep Convolutional Neural Network (DCNN). Various normalization methods have been proposed. The statistics used to normalize the feature maps can be computed at batch, channel, or instance level. However, in most of existing methods, the normalization for each layer is fixed. Batch-Instance Normalization (BIN) is one of the first proposed methods that combines two different normalization methods and achieve diverse normalization for different layers. However, two potential issues exist in BIN: first, the Clip function is not differentiable at input values of 0 and 1; second, the combined feature map is not with a normalized distribution which is harmful for signal propagation in DCNN. In this paper, an Instance-Layer Normalization (ILN) layer is proposed by using the Sigmoid function for the feature map combination, and cascading group normalization. The performance of ILN is validated on image segmentation of the Right Ventricle (RV) and Left Ventricle (LV) using U-Net as the network architecture. The results show that the proposed ILN outperforms previous traditional and popular normalization methods with noticeable accuracy improvements for most validations, supporting the effectiveness of the proposed ILN.
25.Improved MR to CT synthesis for PET/MR attenuation correction using Imitation Learning ⬇️
The ability to synthesise Computed Tomography images - commonly known as pseudo CT, or pCT - from MRI input data is commonly assessed using an intensity-wise similarity, such as an L2-norm between the ground truth CT and the pCT. However, given that the ultimate purpose is often to use the pCT as an attenuation map (
$\mu$ -map) in Positron Emission Tomography Magnetic Resonance Imaging (PET/MRI), minimising the error between pCT and CT is not necessarily optimal. The main objective should be to predict a pCT that, when used as$\mu$ -map, reconstructs a pseudo PET (pPET) which is as close as possible to the gold standard PET. To this end, we propose a novel multi-hypothesis deep learning framework that generates pCTs by minimising a combination of the pixel-wise error between pCT and CT and a proposed metric-loss that itself is represented by a convolutional neural network (CNN) and aims to minimise subsequent PET residuals. The model is trained on a database of 400 paired MR/CT/PET image slices. Quantitative results show that the network generates pCTs that seem less accurate when evaluating the Mean Absolute Error on the pCT (69.68HU) compared to a baseline CNN (66.25HU), but lead to significant improvement in the PET reconstruction - 115a.u. compared to baseline 140a.u.
26.Image Colorization By Capsule Networks ⬇️
In this paper, a simple topology of Capsule Network (CapsNet) is investigated for the problem of image colorization. The generative and segmentation capabilities of the original CapsNet topology, which is proposed for image classification problem, is leveraged for the colorization of the images by modifying the network as follows:1) The original CapsNet model is adapted to map the grayscale input to the output in the CIE Lab colorspace, 2) The feature detector part of the model is updated by using deeper feature layers inherited from VGG-19 pre-trained model with weights in order to transfer low-level image representation capability to this model, 3) The margin loss function is modified as Mean Squared Error (MSE) loss to minimize the image-to-imagemapping. The resulting CapsNet model is named as Colorizer Capsule Network (ColorCapsNet).The performance of the ColorCapsNet is evaluated on the DIV2K dataset and promising results are obtained to investigate Capsule Networks further for image colorization problem.
27.Contour Detection in Cassini ISS images based on Hierarchical Extreme Learning Machine and Dense Conditional Random Field ⬇️
In Cassini ISS (Imaging Science Subsystem) images, contour detection is often performed on disk-resolved object to accurately locate their center. Thus, the contour detection is a key problem. Traditional edge detection methods, such as Canny and Roberts, often extract the contour with too much interior details and noise. Although the deep convolutional neural network has been applied successfully in many image tasks, such as classification and object detection, it needs more time and computer resources. In the paper, a contour detection algorithm based on H-ELM (Hierarchical Extreme Learning Machine) and DenseCRF (Dense Conditional Random Field) is proposed for Cassini ISS images. The experimental results show that this algorithm's performance is better than both traditional machine learning methods such as SVM, ELM and even deep convolutional neural network. And the extracted contour is closer to the actual contour. Moreover, it can be trained and tested quickly on the general configuration of PC, so can be applied to contour detection for Cassini ISS images.
28.Motion correction of dynamic contrast enhanced MRI of the liver ⬇️
Motion correction of dynamic contrast enhanced magnetic resonance images (DCE-MRI) is a challenging task, due to changes in image appearance. In this study a groupwise registration, using a principle component analysis (PCA) based metric,1 is evaluated for clinical DCE MRI of the liver. The groupwise registration transforms the images to a common space, rather than to a reference volume as conventional pairwise methods do, and computes the similarity metric on all volumes simultaneously. This groupwise registration method is compared to a pairwise approach using a mutual information metric. Clinical DCE MRI of the abdomen of eight patients were included. Per patient one lesion in the liver was manually segmented in all temporal images (N=16). The registered images were compared for accuracy, spatial and temporal smoothness after transformation, and lesion volume change. Compared to a pairwise method or no registration, groupwise registration provided better alignment. In our recently started clinical study groupwise registered clinical DCE MRI of the abdomen of nine patients were scored by three radiologists. Groupwise registration increased the assessed quality of alignment. The gain in reading time for the radiologist was estimated to vary from no difference to almost a minute. A slight increase in reader confidence was also observed. Registration had no added value for images with little motion. In conclusion, the groupwise registration of DCE MR images results in better alignment than achieved by pairwise registration, which is beneficial for clinical assessment.
29.Optimal input configuration of dynamic contrast enhanced MRI in convolutional neural networks for liver segmentation ⬇️
Most MRI liver segmentation methods use a structural 3D scan as input, such as a T1 or T2 weighted scan. Segmentation performance may be improved by utilizing both structural and functional information, as contained in dynamic contrast enhanced (DCE) MR series. Dynamic information can be incorporated in a segmentation method based on convolutional neural networks in a number of ways. In this study, the optimal input configuration of DCE MR images for convolutional neural networks (CNNs) is studied. The performance of three different input configurations for CNNs is studied for a liver segmentation task. The three configurations are I) one phase image of the DCE-MR series as input image; II) the separate phases of the DCE-MR as input images; and III) the separate phases of the DCE-MR as channels of one input image. The three input configurations are fed into a dilated fully convolutional network and into a small U-net. The CNNs were trained using 19 annotated DCE-MR series and tested on another 19 annotated DCE-MR series. The performance of the three input configurations for both networks is evaluated against manual annotations. The results show that both neural networks perform better when the separate phases of the DCE-MR series are used as channels of an input image in comparison to one phase as input image or the separate phases as input images. No significant difference between the performances of the two network architectures was found for the separate phases as channels of an input image.
30.NL-LinkNet: Toward Lighter but More Accurate Road Extraction with Non-Local Operations ⬇️
Road extraction from very high resolution satellite images is one of the most important topics in the field of remote sensing. For the road segmentation problem, spatial properties of the data can usually be captured using Convolutional Neural Networks. However, this approach only considers a few local neighborhoods at a time and has difficulty capturing long-range dependencies. In order to overcome the problem, we propose Non-Local LinkNet with non-local blocks that can grasp relations between global features. It enables each spatial feature point to refer to all other contextual information and results in more accurate road segmentation. In detail, our method achieved 65.00% mIOU scores on the DeepGlobe 2018 Road Extraction Challenge dataset. Our best model outperformed D-LinkNet, 1st-ranked solution, by a significant gap of mIOU 0.88% with much less number of parameters. We also present empirical analyses on proper usage of non-local blocks for the baseline model.
31.BIM-assisted object recognition for the on-site autonomous robotic assembly of discrete structures ⬇️
Robots-operating autonomous assembly applications in an unstructured environment require precise methods to locate the building components on site. However, the current available object detection systems are not well-optimised for construction applications, due to the tedious setups incorporated for referencing an object to a system and inability to cope with the elements imperfections. In this paper, we propose a flexible object pose estimation framework to enable robots to autonomously handle building components on-site with an error tolerance to build a specific design target without the need to sort or label them. We implemented an object recognition approach that uses the virtual representation model of all the objects found in a BIM model to autonomously search for the best-matched objects in a scene. The design layout is used to guide the robot to grasp and manipulate the found elements to build the desired structure. We verify our proposed framework by testing it in an automatic discrete wall assembly workflow. Although the precision is not as expected, we analyse the possible reasons that might cause this imprecision, which paves the path for future improvements.
32.An Image Fusion Scheme for Single-Shot High Dynamic Range Imaging with Spatially Varying Exposures ⬇️
This paper proposes a novel multi-exposure image fusion (MEF) scheme for single-shot high dynamic range imaging with spatially varying exposures (SVE). Single-shot imaging with SVE enables us not only to produce images without color saturation regions from a single-shot image, but also to avoid ghost artifacts in the producing ones. However, the number of exposures is generally limited to two, and moreover it is difficult to decide the optimum exposure values before the photographing. In the proposed scheme, a scene segmentation method is applied to input multi-exposure images, and then the luminance of the input images is adjusted according to both of the number of scenes and the relationship between exposure values and pixel values. The proposed method with the luminance adjustment allows us to improve the above two issues. In this paper, we focus on dual-ISO imaging as one of single-shot imaging. In an experiment, the proposed scheme is demonstrated to be effective for single-shot high dynamic range imaging with SVE, compared with conventional MEF schemes with exposure compensation.
33.A CNN toolbox for skin cancer classification ⬇️
We describe a software toolbox for the configuration of deep neural networks in the domain of skin cancer classification. The implemented software architecture allows developers to quickly set up new convolutional neural network (CNN) architectures and hyper-parameter configurations. At the same time, the user interface, manageable as a simple spreadsheet, allows non-technical users to explore different configuration settings that need to be explored when switching to different data sets. In future versions, meta leaning frameworks can be added, or AutoML systems that continuously improve over time. Preliminary results, conducted with two CNNs in the context melanoma detection on dermoscopic images, quantify the impact of image augmentation, image resolution, and rescaling filter on the overall detection performance and training time.
34.Building change detection based on multi-scale filtering and grid partition ⬇️
Building change detection is of great significance in high resolution remote sensing applications. Multi-index learning, one of the state-of-the-art building change detection methods, still has drawbacks like incapability to find change types directly and heavy computation consumption of MBI. In this paper, a two-stage building change detection method is proposed to address these problems. In the first stage, a multi-scale filtering building index (MFBI) is calculated to detect building areas in each temporal with fast speed and moderate accuracy. In the second stage, images and the corresponding building maps are partitioned into grids. In each grid, the ratio of building areas in time T2 and time T1 is calculated. Each grid is classified into one of the three change patterns, i.e., significantly increase, significantly decrease and approximately unchanged. Exhaustive experiments indicate that the proposed method can detect building change types directly and outperform the current multi-index learning method.
35.Transferability and Hardness of Supervised Classification Tasks ⬇️
We propose a novel approach for estimating the difficulty and transferability of supervised classification tasks. Unlike previous work, our approach is solution agnostic and does not require or assume trained models. Instead, we estimate these values using an information theoretic approach: treating training labels as random variables and exploring their statistics. When transferring from a source to a target task, we consider the conditional entropy between two such variables (i.e., label assignments of the two tasks). We show analytically and empirically that this value is related to the loss of the transferred model. We further show how to use this value to estimate task hardness. We test our claims extensively on three large scale data sets -- CelebA (40 tasks), Animals with Attributes 2 (85 tasks), and Caltech-UCSD Birds 200 (312 tasks) -- together representing 437 classification tasks. We provide results showing that our hardness and transferability estimates are strongly correlated with empirical hardness and transferability. As a case study, we transfer a learned face recognition model to CelebA attribute classification tasks, showing state of the art accuracy for tasks estimated to be highly transferable.
36.DUAL-GLOW: Conditional Flow-Based Generative Model for Modality Transfer ⬇️
Positron emission tomography (PET) imaging is an imaging modality for diagnosing a number of neurological diseases. In contrast to Magnetic Resonance Imaging (MRI), PET is costly and involves injecting a radioactive substance into the patient. Motivated by developments in modality transfer in vision, we study the generation of certain types of PET images from MRI data. We derive new flow-based generative models which we show perform well in this small sample size regime (much smaller than dataset sizes available in standard vision tasks). Our formulation, DUAL-GLOW, is based on two invertible networks and a relation network that maps the latent spaces to each other. We discuss how given the prior distribution, learning the conditional distribution of PET given the MRI image reduces to obtaining the conditional distribution between the two latent codes w.r.t. the two image types. We also extend our framework to leverage 'side' information (or attributes) when available. By controlling the PET generation through 'conditioning' on age, our model is also able to capture brain FDG-PET (hypometabolism) changes, as a function of age. We present experiments on the Alzheimers Disease Neuroimaging Initiative (ADNI) dataset with 826 subjects, and obtain good performance in PET image synthesis, qualitatively and quantitatively better than recent works.
37.Boundary Aware Networks for Medical Image Segmentation ⬇️
Fully convolutional neural networks (CNNs) have proven to be effective at representing and classifying textural information, thus transforming image intensity into output class masks that achieve semantic image segmentation. In medical image analysis, however, expert manual segmentation often relies on the boundaries of anatomical structures of interest. We propose boundary aware CNNs for medical image segmentation. Our networks are designed to account for organ boundary information, both by providing a special network edge branch and edge-aware loss terms, and they are trainable end-to-end. We validate their effectiveness on the task of brain tumor segmentation using the BraTS 2018 dataset. Our experiments reveal that our approach yields more accurate segmentation results, which makes it promising for more extensive application to medical image segmentation.
38.More unlabelled data or label more data? A study on semi-supervised laparoscopic image segmentation ⬇️
Improving a semi-supervised image segmentation task has the option of adding more unlabelled images, labelling the unlabelled images or combining both, as neither image acquisition nor expert labelling can be considered trivial in most clinical applications. With a laparoscopic liver image segmentation application, we investigate the performance impact by altering the quantities of labelled and unlabelled training data, using a semi-supervised segmentation algorithm based on the mean teacher learning paradigm. We first report a significantly higher segmentation accuracy, compared with supervised learning. Interestingly, this comparison reveals that the training strategy adopted in the semi-supervised algorithm is also responsible for this observed improvement, in addition to the added unlabelled data. We then compare different combinations of labelled and unlabelled data set sizes for training semi-supervised segmentation networks, to provide a quantitative example of the practically useful trade-off between the two data planning strategies in this surgical guidance application.
39.InSituNet: Deep Image Synthesis for Parameter Space Exploration of Ensemble Simulations ⬇️
We propose InSituNet, a deep learning based surrogate model to support parameter space exploration for ensemble simulations that are visualized in situ. In situ visualization, generating visualizations at simulation time, is becoming prevalent in handling large-scale simulations because of the I/O and storage constraints. However, in situ visualization approaches limit the flexibility of post-hoc exploration because the raw simulation data are no longer available. Although multiple image-based approaches have been proposed to mitigate this limitation, those approaches lack the ability to explore the simulation parameters. Our approach allows flexible exploration of parameter space for large-scale ensemble simulations by taking advantage of the recent advances in deep learning. Specifically, we design InSituNet as a convolutional regression model to learn the mapping from the simulation and visualization parameters to the visualization results. With the trained model, users can generate new images for different simulation parameters under various visualization settings, which enables in-depth analysis of the underlying ensemble simulations. We demonstrate the effectiveness of InSituNet in combustion, cosmology, and ocean simulations through quantitative and qualitative evaluations.