1.Splenomegaly Segmentation on Multi-modal MRI using Deep Convolutional Networks pdf
The findings of splenomegaly, abnormal enlargement of the spleen, is a non-invasive clinical biomarker for liver and spleen disease. Automated segmentation methods are essential to efficiently quantify splenomegaly from clinically acquired abdominal magnetic resonance imaging (MRI) scans. However, the task is challenging due to (1) large anatomical and spatial variations of splenomegaly, (2) large inter- and intra-scan intensity variations on multi-modal MRI, and (3) limited numbers of labeled splenomegaly scans. In this paper, we propose the Splenomegaly Segmentation Network (SS-Net) to introduce the deep convolutional neural network (DCNN) approaches in multi-modal MRI splenomegaly segmentation. Large convolutional kernel layers were used to address the spatial and anatomical variations, while the conditional generative adversarial networks (GAN) were employed to leverage the segmentation performance of SS-Net in an end-to-end manner. A clinically acquired cohort containing both T1-weighted (T1w) and T2-weighted (T2w) MRI splenomegaly scans was used to train and evaluate the performance of multi-atlas segmentation (MAS), 2D DCNN networks, and a 3D DCNN network. From the experimental results, the DCNN methods achieved superior performance to the state-of-the-art MAS method. The proposed SS-Net method achieved the highest median and mean Dice scores among investigated baseline DCNN methods.
2.Identify, locate and separate: Audio-visual object extraction in large video collections using weak supervision pdf
We tackle the problem of audiovisual scene analysis for weakly-labeled data. To this end, we build upon our previous audiovisual representation learning framework to perform object classification in noisy acoustic environments and integrate audio source enhancement capability. This is made possible by a novel use of non-negative matrix factorization for the audio modality. Our approach is founded on the multiple instance learning paradigm. Its effectiveness is established through experiments over a challenging dataset of music instrument performance videos. We also show encouraging visual object localization results.
3.Matrix Recovery with Implicitly Low-Rank Data pdf
In this paper, we study the problem of matrix recovery, which aims to restore a target matrix of authentic samples from grossly corrupted observations. Most of the existing methods, such as the well-known Robust Principal Component Analysis (RPCA), assume that the target matrix we wish to recover is low-rank. However, the underlying data structure is often non-linear in practice, therefore the low-rankness assumption could be violated. To tackle this issue, we propose a novel method for matrix recovery in this paper, which could well handle the case where the target matrix is low-rank in an implicit feature space but high-rank or even full-rank in its original form. Namely, our method pursues the low-rank structure of the target matrix in an implicit feature space. By making use of the specifics of an accelerated proximal gradient based optimization algorithm, the proposed method could recover the target matrix with non-linear structures from its corrupted version. Comprehensive experiments on both synthetic and real datasets demonstrate the superiority of our method.
4.Cross and Learn: Cross-Modal Self-Supervision pdf
In this paper we present a self-supervised method for representation learning utilizing two different modalities. Based on the observation that cross-modal information has a high semantic meaning we propose a method to effectively exploit this signal. For our approach we utilize video data since it is available on a large scale and provides easily accessible modalities given by RGB and optical flow. We demonstrate state-of-the-art performance on highly contested action recognition datasets in the context of self-supervised learning. We show that our feature representation also transfers to other tasks and conduct extensive ablation studies to validate our core contributions.
5.An Average of the Human Ear Canal: Recovering Acoustical Properties via Shape Analysis pdf
Humans are highly dependent on the ability to process audio in order to interact through conversation and navigate from sound. For this, the shape of the ear acts as a mechanical audio filter. The anatomy of the outer human ear canal to approximately 15-20 mm beyond the Tragus is well described because of its importance for customized hearing aid production. This is however not the case for the part of the ear canal that is embedded in the skull, until the typanic membrane. Due to the sensitivity of the outer ear, this part, referred to as the bony part, has only been described in a few population studies and only ex-vivo. We present a study of the entire ear canal including the bony part and the tympanic membrane. We form an average ear canal from a number of MRI scans using standard image registration methods. We show that the obtained representation is realistic in the sense that it has acoustical properties almost identical to a real ear.
6.Image-Level Attentional Context Modeling Using Nested-Graph Neural Networks pdf
We introduce a new scene graph generation method called image-level attentional context modeling (ILAC). Our model includes an attentional graph network that effectively propagates contextual information across the graph using image-level features. Whereas previous works use an object-centric context, we build an image-level context agent to encode the scene properties. The proposed method comprises a single-stream network that iteratively refines the scene graph with a nested graph neural network. We demonstrate that our approach achieves competitive performance with the state-of-the-art for scene graph generation on the Visual Genome dataset, while requiring fewer parameters than other methods. We also show that ILAC can improve regular object detectors by incorporating relational image-level information.
7.Changing the Image Memorability: From Basic Photo Editing to GANs pdf
Memorability is considered to be an important characteristic of visual content, whereas for advertisement and educational purposes it is the most important one. Despite numerous studies on understanding and predicting image memorability, there are almost no achievements in memorability modification. In this work, we study two possible approaches to image modification which likely may influence memorability. The visual features which influence memorability directly stay unknown till now, hence it is impossible to control it manually. As a solution, we let GAN learn it deeply using labeled data, and then use it for conditional generation of new images. By analogy with algorithms which edit facial attributes, we consider memorability as yet another attribute and operate with it in the same way. Obtained data is also interesting for analysis, simply because there are no real-world examples of successful change of image memorability while preserving its other attributes. We believe this may give many new answers to the question "what makes an image memorable?" Apart from that we also study the influence of conventional photo-editing tools (Photoshop, Instagram, etc.) used daily by a wide audience on memorability. In this case, we start from real practical methods and study it using statistics and recent advances in memorability prediction. Photographers, designers, and advertisers will benefit from the results of this study directly.
8.RoarNet: A Robust 3D Object Detection based on RegiOn Approximation Refinement pdf
We present RoarNet, a new approach for 3D object detection from a 2D image and 3D Lidar point clouds. Based on two-stage object detection framework with PointNet as our backbone network, we suggest several novel ideas to improve 3D object detection performance. The first part of our method, RoarNet_2D, estimates the 3D poses of objects from a monocular image, which approximates where to examine further, and derives multiple candidates that are geometrically feasible. This step significantly narrows down feasible 3D regions, which otherwise requires demanding processing of 3D point clouds in a huge search space. Then the second part, RoarNet_3D, takes the candidate regions and conducts in-depth inferences to conclude final poses in a recursive manner. Inspired by PointNet, RoarNet_3D processes 3D point clouds directly without any loss of data, leading to precise detection. We evaluate our method in KITTI, a 3D object detection benchmark. Our result shows that RoarNet has superior performance to state-of-the-art methods that are publicly available. Remarkably, RoarNet also outperforms state-of-the-art methods even in settings where Lidar and camera are not time synchronized, which is practically important for actual driving environments. RoarNet is implemented in Tensorflow and publicly available with pre-trained models.
9.Neural Stain Normalization and Unsupervised Classification of Cell Nuclei in Histopathological Breast Cancer Images pdf
In this paper, we develop a complete pipeline for stain normalization, segmentation, and classification of nuclei in hematoxylin and eosin (H&E) stained breast cancer histopathology images. In the first step, we use a CNN-based stain transfer technique to normalize the staining characteristics of (H&E) images. We then train a neural network to segment images of nuclei from the H&E images. Finally, we train an Information Maximizing Generative Adversarial Network (InfoGAN) to learn visual representations of different types of nuclei and classify them in an entirely unsupervised manner. The results show that our proposed CNN stain normalization yields improved visual similarity and cell segmentation performance compared to the conventional SVD-based stain normalization method. In the final step of our pipeline, we demonstrate the ability to perform fully unsupervised clustering of various breast histopathology cell types based on morphological and color attributes. In addition, we quantitatively evaluate our neural network - based techniques against various quantitative metrics to validate the effectiveness of our pipeline.
10.A Theoretically Guaranteed Deep Optimization Framework for Robust Compressive Sensing MRI pdf
Magnetic Resonance Imaging (MRI) is one of the most dynamic and safe imaging techniques available for clinical applications. However, the rather slow speed of MRI acquisitions limits the patient throughput and potential indi cations. Compressive Sensing (CS) has proven to be an efficient technique for accelerating MRI acquisition. The most widely used CS-MRI model, founded on the premise of reconstructing an image from an incompletely filled k-space, leads to an ill-posed inverse problem. In the past years, lots of efforts have been made to efficiently optimize the CS-MRI model. Inspired by deep learning techniques, some preliminary works have tried to incorporate deep architectures into CS-MRI process. Unfortunately, the convergence issues (due to the experience-based networks) and the robustness (i.e., lack real-world noise modeling) of these deeply trained optimization methods are still missing. In this work, we develop a new paradigm to integrate designed numerical solvers and the data-driven architectures for CS-MRI. By introducing an optimal condition checking mechanism, we can successfully prove the convergence of our established deep CS-MRI optimization scheme. Furthermore, we explicitly formulate the Rician noise distributions within our framework and obtain an extended CS-MRI network to handle the real-world nosies in the MRI process. Extensive experimental results verify that the proposed paradigm outperforms the existing state-of-the-art techniques both in reconstruction accuracy and efficiency as well as robustness to noises in real scene.
11.A Fully Automated System for Sizing Nasal PAP Masks Using Facial Photographs pdf
We present a fully automated system for sizing nasal Positive Airway Pressure (PAP) masks. The system is comprised of a mix of HOG object detectors as well as multiple convolutional neural network stages for facial landmark detection. The models were trained using samples from the publicly available PUT and MUCT datasets while transfer learning was also employed to improve the performance of the models on facial photographs of actual PAP mask users. The fully automated system demonstrated an overall accuracy of 64.71% in correctly selecting the appropriate mask size and 86.1% accuracy sizing within 1 mask size.
12.M2M-GAN: Many-to-Many Generative Adversarial Transfer Learning for Person Re-Identification pdf
Cross-domain transfer learning (CDTL) is an extremely challenging task for the person re-identification (ReID). Given a source domain with annotations and a target domain without annotations, CDTL seeks an effective method to transfer the knowledge from the source domain to the target domain. However, such a simple two-domain transfer learning method is unavailable for the person ReID in that the source/target domain consists of several sub-domains, e.g., camera-based sub-domains. To address this intractable problem, we propose a novel Many-to-Many Generative Adversarial Transfer Learning method (M2M-GAN) that takes multiple source sub-domains and multiple target sub-domains into consideration and performs each sub-domain transferring mapping from the source domain to the target domain in a unified optimization process. The proposed method first translates the image styles of source sub-domains into that of target sub-domains, and then performs the supervised learning by using the transferred images and the corresponding annotations in source domain. As the gap is reduced, M2M-GAN achieves a promising result for the cross-domain person ReID. Experimental results on three benchmark datasets Market-1501, DukeMTMC-reID and MSMT17 show the effectiveness of our M2M-GAN.
13.Typeface Completion with Generative Adversarial Networks pdf
The mood of a text and the intention of the writer can be reflected in the typeface. However, in designing a typeface, it is difficult to keep the style of various characters consistent, especially for languages with lots of morphological variations such as Chinese. In this paper, we propose a Typeface Completion Network (TCN) which takes a subset of characters as an input, and automatically completes the entire set of characters in the same style as the input characters. Unlike existing models proposed for style transfer, TCN embeds a character image into two separate vectors representing typeface and content. Combined with a reconstruction loss from the latent space, and with other various losses, TCN overcomes the inherent difficulty in designing a typeface. Also, compared to previous style transfer models, TCN generates high quality characters of the same typeface with a much smaller number of model parameters. We validate our proposed model on the Chinese and English character datasets, and the CelebA dataset on which TCN outperforms recently proposed state-ofthe-art models for style transfer. The source code of our model is available at this https URL.
14.Semantic and Contrast-Aware Saliency pdf
In this paper, we proposed an integrated model of semantic-aware and contrast-aware saliency combining both bottom-up and top-down cues for effective saliency estimation and eye fixation prediction. The proposed model processes visual information using two pathways. The first pathway aims to capture the attractive semantic information in images, especially for the presence of meaningful objects and object parts such as human faces. The second pathway is based on multi-scale on-line feature learning and information maximization, which learns an adaptive sparse representation for the input and discovers the high contrast salient patterns within the image context. The two pathways characterize both long-term and short-term attention cues and are integrated dynamically using maxima normalization. We investigate two different implementations of the semantic pathway including an End-to-End deep neural network solution and a dynamic feature integration solution, resulting in the SCA and SCAFI model respectively. Experimental results on artificial images and 5 popular benchmark datasets demonstrate the superior performance and better plausibility of the proposed model over both classic approaches and recent deep models.
15.Learning Energy Based Inpainting for Optical Flow pdf
Modern optical flow methods are often composed of a cascade of many independent steps or formulated as a black box neural network that is hard to interpret and analyze. In this work we seek for a plain, interpretable, but learnable solution. We propose a novel inpainting based algorithm that approaches the problem in three steps: feature selection and matching, selection of supporting points and energy based inpainting. To facilitate the inference we propose an optimization layer that allows to backpropagate through 10K iterations of a first-order method without any numerical or memory problems. Compared to recent state-of-the-art networks, our modular CNN is very lightweight and competitive with other, more involved, inpainting based methods.
16.Validating Hyperspectral Image Segmentation pdf
Hyperspectral satellite imaging attracts enormous research attention in the remote sensing community, hence automated approaches for precise segmentation of such imagery are being rapidly developed. In this letter, we share our observations on the strategy for validating hyperspectral image segmentation algorithms currently followed in the literature, and show that it can lead to over-optimistic experimental insights. We introduce a new routine for generating segmentation benchmarks, and use it to elaborate ready-to-use hyperspectral training-test data partitions. They can be utilized for fair validation of new and existing algorithms without any training-test data leakage.
17.Deep Learning Predicts Hip Fracture using Confounding Patient and Healthcare Variables pdf
Hip fractures are a leading cause of death and disability among older adults. Hip fractures are also the most commonly missed diagnosis on pelvic radiographs. Computer-Aided Diagnosis (CAD) algorithms have shown promise for helping radiologists detect fractures, but the image features underpinning their predictions are notoriously difficult to understand. In this study, we trained deep learning models on 17,587 radiographs to classify fracture, five patient traits, and 14 hospital process variables. All 20 variables could be predicted from a radiograph (p < 0.05), with the best performances on scanner model (AUC=1.00), scanner brand (AUC=0.98), and whether the order was marked "priority" (AUC=0.79). Fracture was predicted moderately well from the image (AUC=0.78) and better when combining image features with patient data (AUC=0.86, p=2e-9) or patient data plus hospital process features (AUC=0.91, p=1e-21). The model performance on a test set with matched patient variables was significantly lower than a random test set (AUC=0.67, p=0.003); and when the test set was matched on patient and image acquisition variables, the model performed randomly (AUC=0.52, 95% CI 0.46-0.58), indicating that these variables were the main source of the model's predictive ability overall. We also used Naive Bayes to combine evidence from image models with patient and hospital data and found their inclusion improved performance, but that this approach was nevertheless inferior to directly modeling all variables. If CAD algorithms are inexplicably leveraging patient and process variables in their predictions, it is unclear how radiologists should interpret their predictions in the context of other known patient data. Further research is needed to illuminate deep learning decision processes so that computers and clinicians can effectively cooperate.
18.NEMGAN: Noise Engineered Mode-matching GAN pdf
Conditional generation refers to the process of sampling from an unknown distribution conditioned on semantics of the data. This can be achieved by augmenting the generative model with the desired semantic labels, albeit it is not straightforward in an unsupervised setting where the semantic label of every data sample is unknown. In this paper, we address this issue by proposing a method that can generate samples conditioned on the properties of a latent distribution engineered in accordance with a certain data prior. In particular, a latent space inversion network is trained in tandem with a generative adversarial network such that the modal properties of the latent space distribution are induced in the data generating distribution. We demonstrate that our model despite being fully unsupervised, is effective in learning meaningful representations through its mode matching property. We validate our method on multiple unsupervised tasks such as conditional generation, attribute discovery and inference using three real world image datasets namely MNIST, CIFAR-10 and CelebA and show that the results are comparable to the state-of-the-art methods.
19.Can Deep Learning Outperform Modern Commercial CT Image Reconstruction Methods? pdf
Commercial iterative reconstruction techniques on modern CT scanners target radiation dose reduction but there are lingering concerns over their impact on image appearance and low contrast detectability. Recently, machine learning, especially deep learning, has been actively investigated for CT. Here we design a novel neural network architecture for low-dose CT (LDCT) and compare it with commercial iterative reconstruction methods used for standard of care CT. While popular neural networks are trained for end-to-end mapping, driven by big data, our novel neural network is intended for end-to-process mapping so that intermediate image targets are obtained with the associated search gradients along which the final image targets are gradually reached. This learned dynamic process allows to include radiologists in the training loop to optimize the LDCT denoising workflow in a task-specific fashion with the denoising depth as a key parameter. Our progressive denoising network was trained with the Mayo LDCT Challenge Dataset, and tested on images of the chest and abdominal regions scanned on the CT scanners made by three leading CT vendors. The best deep learning based reconstructions are systematically compared to the best iterative reconstructions in a double-blinded reader study. It is found that our deep learning approach performs either comparably or favorably in terms of noise suppression and structural fidelity, and runs orders of magnitude faster than the commercial iterative CT reconstruction algorithms.
20.Gender Effect on Face Recognition for a Large Longitudinal Database pdf
Aging or gender variation can affect the face recognition performance dramatically. While most of the face recognition studies are focused on the variation of pose, illumination and expression, it is important to consider the influence of gender effect and how to design an effective matching framework. In this paper, we address these problems on a very large longitudinal database MORPH-II which contains 55,134 face images of 13,617 individuals. First, we consider four comprehensive experiments with different combination of gender distribution and subset size, including: 1) equal gender distribution; 2) a large highly unbalanced gender distribution; 3) consider different gender combinations, such as male only, female only, or mixed gender; and 4) the effect of subset size in terms of number of individuals. Second, we consider eight nearest neighbor distance metrics and also Support Vector Machine (SVM) for classifiers and test the effect of different classifiers. Last, we consider different fusion techniques for an effective matching framework to improve the recognition performance.
21.Evolvement Constrained Adversarial Learning for Video Style Transfer pdf
Video style transfer is a useful component for applications such as augmented reality, non-photorealistic rendering, and interactive games. Many existing methods use optical flow to preserve the temporal smoothness of the synthesized video. However, the estimation of optical flow is sensitive to occlusions and rapid motions. Thus, in this work, we introduce a novel evolve-sync loss computed by evolvements to replace optical flow. Using this evolve-sync loss, we build an adversarial learning framework, termed as Video Style Transfer Generative Adversarial Network (VST-GAN), which improves upon the MGAN method for image style transfer for more efficient video style transfer. We perform extensive experimental evaluations of our method and show quantitative and qualitative improvements over the state-of-the-art methods.
22.Convolutional LSTMs for Cloud-Robust Segmentation of Remote Sensing Imagery pdf
Dynamic spatiotemporal processes on the Earth can be observed by an increasing number of optical Earth observation satellites that measure spectral reflectance at multiple spectral bands in regular intervals. Clouds partially covering the surface is an omnipresent challenge for the majority of remote sensing approaches that are not robust regarding cloud coverage. In these approaches, clouds are typically handled by cherry-picking cloud-free observations or by pre-classification of cloudy pixels and subsequent masking. In this work, we demonstrate the robustness of a straightforward convolutional long short-term memory network for vegetation classification using all available cloudy and non-cloudy satellite observations. We visualize the internal gate activations within the recurrent cells and find that, in some cells, modulation and input gates close on cloudy pixels. This indicates that the network has internalized a cloud-filtering mechanism without being specifically trained on cloud labels. The robustness regarding clouds is further demonstrated by experiments on sequences with varying degrees of cloud coverage where our network achieved similar accuracies on all cloudy and non-cloudy datasets. Overall, our results question the necessity of sophisticated pre-processing pipelines if robust classification methods are utilized.
23.SparseFool: a few pixels make a big difference pdf
Deep Neural Networks have achieved extraordinary results on image classification tasks, but have been shown to be vulnerable to attacks with carefully crafted perturbations of the input data. Although most attacks usually change values of many image's pixels, it has been shown that deep networks are also vulnerable to sparse alterations of the input. However, no efficient method has been proposed to compute sparse perturbations. In this paper, we exploit the low mean curvature of the decision boundary, and propose SparseFool, a geometry inspired sparse attack that controls the sparsity of the perturbations. Extensive evaluations show that our approach outperforms related methods, and scales to high dimensional data. We further analyze the transferability and the visual effects of the perturbations, and show the existence of shared semantic information across the images and the networks. Finally, we show that adversarial training using
$\ell_\infty$ perturbations can slightly improve the robustness against sparse additive perturbations.
24.Weakly Supervised Scene Parsing with Point-based Distance Metric Learning pdf
Semantic scene parsing is suffering from the fact that pixel-level annotations are hard to be collected. To tackle this issue, we propose a Point-based Distance Metric Learning (PDML) in this paper. PDML does not require dense annotated masks and only leverages several labeled points that are much easier to obtain to guide the training process. Concretely, we leverage semantic relationship among the annotated points by encouraging the feature representations of the intra- and inter-category points to keep consistent, i.e. points within the same category should have more similar feature representations compared to those from different categories. We formulate such a characteristic into a simple distance metric loss, which collaborates with the point-wise cross-entropy loss to optimize the deep neural networks. Furthermore, to fully exploit the limited annotations, distance metric learning is conducted across different training images instead of simply adopting an image-dependent manner. We conduct extensive experiments on two challenging scene parsing benchmarks of PASCAL-Context and ADE 20K to validate the effectiveness of our PDML, and competitive mIoU scores are achieved.
25.BLP - Boundary Likelihood Pinpointing Networks for Accurate Temporal Action Localization pdf
Despite tremendous progress achieved in temporal action detection, state-of-the-art methods still suffer from the sharp performance deterioration when localizing the starting and ending temporal action boundaries. Although most methods apply boundary regression paradigm to tackle this problem, we argue that the direct regression lacks detailed enough information to yield accurate temporal boundaries. In this paper, we propose a novel Boundary Likelihood Pinpointing (BLP) network to alleviate this deficiency of boundary regression and improve the localization accuracy. Given a loosely localized search interval that contains an action instance, BLP casts the problem of localizing temporal boundaries as that of assigning probabilities on each equally divided unit of this interval. These generated probabilities provide useful information regarding the boundary location of the action inside this search interval. Based on these probabilities, we introduce a boundary pinpointing paradigm to pinpoint the accurate boundaries under a simple probabilistic framework. Compared with other C3D feature based detectors, extensively experiments demonstrate that BLP significantly improve the localization performance of recent state-of-the-art detectors, and achieve competitive detection mAP on both THUMOS' 14 and ActivityNet datasets, particularly when the evaluation tIoU is high.
26.A Microprocessor implemented in 65nm CMOS with Configurable and Bit-scalable Accelerator for Programmable In-memory Computing pdf
This paper presents a programmable in-memory-computing processor, demonstrated in a 65nm CMOS technology. For data-centric workloads, such as deep neural networks, data movement often dominates when implemented with today's computing architectures. This has motivated spatial architectures, where the arrangement of data-storage and compute hardware is distributed and explicitly aligned to the computation dataflow, most notably for matrix-vector multiplication. In-memory computing is a spatial architecture where processing elements correspond to dense bit cells, providing local storage and compute, typically employing analog operation. Though this raises the potential for high energy efficiency and throughput, analog operation has significantly limited robustness, scale, and programmability. This paper describes a 590kb in-memory-computing accelerator integrated in a programmable processor architecture, by exploiting recent approaches to charge-domain in-memory computing. The architecture takes the approach of tight coupling with an embedded CPU, through accelerator interfaces enabling integration in the standard processor memory space. Additionally, a near-memory-computing datapath both enables diverse computations locally, to address operations required across applications, and enables bit-precision scalability for matrix/input-vector elements, through a bit-parallel/bit-serial (BP/BS) scheme. Chip measurements show an energy efficiency of 152/297 1b-TOPS/W and throughput of 4.7/1.9 1b-TOPS (scaling linearly with the matrix/input-vector element precisions) at VDD of 1.2/0.85V. Neural network demonstrations with 1-b/4-b weights and activations for CIFAR-10 classification consume 5.3/105.2 $μ$J/image at 176/23 fps, with accuracy at the level of digital/software implementation (89.3/92.4
$%$ accuracy).
27.Toward Autonomous Rotation-Aware Unmanned Aerial Grasping pdf
Autonomous Unmanned Aerial Manipulators (UAMs) have shown promising potentials to transform passive sensing missions into active 3-dimension interactive missions, but they still suffer from some difficulties impeding their wide applications, such as target detection and stabilization. This letter presents a vision-based autonomous UAM with a 3DoF robotic arm for rotational grasping, with a compensation on displacement for center of gravity. First, the hardware, software architecture and state estimation methods are detailed. All the mechanical designs are fully provided as open-source hardware for the reuse by the community. Then, we analyze the flow distribution generated by rotors and plan the robotic arm's motion based on this analysis. Next, a novel detection approach called Rotation-SqueezeDet is proposed to enable rotation-aware grasping, which can give the target position and rotation angle in near real-time on Jetson TX2. Finally, the effectiveness of the proposed scheme is validated in multiple experimental trials, highlighting it's applicability of autonomous aerial grasping in GPS-denied environments.
28.Multimodal One-Shot Learning of Speech and Images pdf
Imagine a robot is shown new concepts visually together with spoken tags, e.g. "milk", "eggs", "butter". After seeing one paired audio-visual example per class, it is shown a new set of unseen instances of these objects, and asked to pick the "milk". Without receiving any hard labels, could it learn to match the new continuous speech input to the correct visual instance? Although unimodal one-shot learning has been studied, where one labelled example in a single modality is given per class, this example motivates multimodal one-shot learning. Our main contribution is to formally define this task, and to propose several baseline and advanced models. We use a dataset of paired spoken and visual digits to specifically investigate recent advances in Siamese convolutional neural networks. Our best Siamese model achieves twice the accuracy of a nearest neighbour model using pixel-distance over images and dynamic time warping over speech in 11-way cross-modal matching.
29.Gradient Descent Finds Global Minima of Deep Neural Networks pdf
Gradient descent finds a global minimum in training deep neural networks despite the objective function being non-convex. The current paper proves gradient descent achieves zero training loss in polynomial time for a deep over-parameterized neural network with residual connections (ResNet). Our analysis relies on the particular structure of the Gram matrix induced by the neural network architecture. This structure allows us to show the Gram matrix is stable throughout the training process and this stability implies the global optimality of the gradient descent algorithm. Our bounds also shed light on the advantage of using ResNet over the fully connected feedforward architecture; our bound requires the number of neurons per layer scaling exponentially with depth for feedforward networks whereas for ResNet the bound only requires the number of neurons per layer scaling polynomially with depth. We further extend our analysis to deep residual convolutional neural networks and obtain a similar convergence result.
30.Satyam: Democratizing Groundtruth for Machine Vision pdf
The democratization of machine learning (ML) has led to ML-based machine vision systems for autonomous driving, traffic monitoring, and video surveillance. However, true democratization cannot be achieved without greatly simplifying the process of collecting groundtruth for training and testing these systems. This groundtruth collection is necessary to ensure good performance under varying conditions. In this paper, we present the design and evaluation of Satyam, a first-of-its-kind system that enables a layperson to launch groundtruth collection tasks for machine vision with minimal effort. Satyam leverages a crowdtasking platform, Amazon Mechanical Turk, and automates several challenging aspects of groundtruth collection: creating and launching of custom web-UI tasks for obtaining the desired groundtruth, controlling result quality in the face of spammers and untrained workers, adapting prices to match task complexity, filtering spammers and workers with poor performance, and processing worker payments. We validate Satyam using several popular benchmark vision datasets, and demonstrate that groundtruth obtained by Satyam is comparable to that obtained from trained experts and provides matching ML performance when used for training.
31.Synaptic Strength For Convolutional Neural Network pdf
Convolutional Neural Networks(CNNs) are both computation and memory intensive which hindered their deployment in mobile devices. Inspired by the relevant concept in neural science literature, we propose Synaptic Pruning: a data-driven method to prune connections between input and output feature maps with a newly proposed class of parameters called Synaptic Strength. Synaptic Strength is designed to capture the importance of a connection based on the amount of information it transports. Experiment results show the effectiveness of our approach. On CIFAR-10, we prune connections for various CNN models with up to 96% , which results in significant size reduction and computation saving. Further evaluation on ImageNet demonstrates that synaptic pruning is able to discover efficient models which is competitive to state-of-the-art compact CNNs such as MobileNet-V2 and NasNet-Mobile. Our contribution is summarized as following: (1) We introduce Synaptic Strength, a new class of parameters for CNNs to indicate the importance of each connections. (2) Our approach can prune various CNNs with high compression without compromising accuracy. (3) Further investigation shows, the proposed Synaptic Strength is a better indicator for kernel pruning compared with the previous approach in both empirical result and theoretical analysis.