PyTorchKR Papers

최신 논문

총 2,147편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1707.09700공식 코드 · Papers with Code

Scene Graph Generation from Objects, Phrases and Region Captions

Object detection, scene graph generation and region captioning, which are three scene understanding tasks at different semantic levels, are tied together: scene graphs are generated on top of objects detected in an image with their pairwise relationship predicted, while region captioning gives a language description of the objects, their attributes, relations, and other context information. In this work, to leverage the mutual connections across semantic levels, we propose a novel neural network model, termed as Multi-level Scene Description Network (denoted as MSDN), to solve the three vision tasks jointly in an end-to-end manner. Objects, phrases, and caption regions are first aligned with a dynamic graph based on their spatial and semantic connections. Then a feature refining structure is used to pass messages across the three levels of semantic tasks through the graph. We benchmark the learned model on three tasks, and show the joint learning across three tasks with our proposed method can bring mutual improvements over previous models. Particularly, on the scene graph generation task, our proposed method outperforms the state-of-art method with more than 3% margin.

arXiv:1709.05774

Direction-Aware Semi-Dense SLAM

To aide simultaneous localization and mapping (SLAM), future perception systems will incorporate forms of scene understanding. In a step towards fully integrated probabilistic geometric scene understanding, localization and mapping we propose the first direction-aware semi-dense SLAM system. It jointly infers the directional Stata Center World (SCW) segmentation and a surfel-based semi-dense map while performing real-time camera tracking. The joint SCW map model connects a scene-wide Bayesian nonparametric Dirichlet Process von-Mises-Fisher mixture model (DP-vMF) prior on surfel orientations with the local surfel locations via a conditional random field (CRF). Camera tracking leverages the SCW segmentation to improve efficiency via guided observation selection. Results demonstrate improved SLAM accuracy and tracking efficiency at state of the art performance.

arXiv:1709.06158공식 코드 · Papers with Code

Matterport3D: Learning from RGB-D Data in Indoor Environments

Access to large, diverse RGB-D datasets is critical for training RGB-D scene understanding algorithms. However, existing datasets still cover only a limited number of views or a restricted scale of spaces. In this paper, we introduce Matterport3D, a large-scale RGB-D dataset containing 10,800 panoramic views from 194,400 RGB-D images of 90 building-scale scenes. Annotations are provided with surface reconstructions, camera poses, and 2D and 3D semantic segmentations. The precise global alignment and comprehensive, diverse panoramic set of views over entire buildings enable a variety of supervised and self-supervised computer vision tasks, including keypoint matching, view overlap prediction, normal prediction from color, semantic segmentation, and region classification.

arXiv:1710.00132

Dense RGB-D semantic mapping with Pixel-Voxel neural network

For intelligent robotics applications, extending 3D mapping to 3D semantic mapping enables robots to, not only localize themselves with respect to the scene's geometrical features but also simultaneously understand the higher level meaning of the scene contexts. Most previous methods focus on geometric 3D reconstruction and scene understanding independently notwithstanding the fact that joint estimation can boost the accuracy of the semantic mapping. In this paper, a dense RGB-D semantic mapping system with a Pixel-Voxel network is proposed, which can perform dense 3D mapping while simultaneously recognizing and semantically labelling each point in the 3D map. The proposed Pixel-Voxel network obtains global context information by using PixelNet to exploit the RGB image and meanwhile, preserves accurate local shape information by using VoxelNet to exploit the corresponding 3D point cloud. Unlike the existing architecture that fuses score maps from different models with equal weights, we proposed a Softmax weighted fusion stack that adaptively learns the varying contributions of PixelNet and VoxelNet, and fuses the score maps of the two models according to their respective confidence levels. The proposed Pixel-Voxel network achieves the state-of-the-art semantic segmentation performance on the SUN RGB-D benchmark dataset. The runtime of the proposed system can be boosted to 11-12Hz, enabling near to real-time performance using an i7 8-cores PC with Titan X GPU.

arXiv:1711.06354

Grounded Objects and Interactions for Video Captioning

We address the problem of video captioning by grounding language generation on object interactions in the video. Existing work mostly focuses on overall scene understanding with often limited or no emphasis on object interactions to address the problem of video understanding. In this paper, we propose SINet-Caption that learns to generate captions grounded over higher-order interactions between arbitrary groups of objects for fine-grained video understanding. We discuss the challenges and benefits of such an approach. We further demonstrate state-of-the-art results on the ActivityNet Captions dataset using our model, SINet-Caption based on this approach.

arXiv:1609.05834

On Support Relations and Semantic Scene Graphs

Scene understanding is a popular and challenging topic in both computer vision and photogrammetry. Scene graph provides rich information for such scene understanding. This paper presents a novel approach to infer such relations and then to construct the scene graph. Support relations are estimated by considering important, previously ignored information: the physical stability and the prior support knowledge between object classes. In contrast to previous methods for extracting support relations, the proposed approach generates more accurate results, and does not require a pixel-wise semantic labeling of the scene. The semantic scene graph which describes all the contextual relations within the scene is constructed using this information. To evaluate the accuracy of these graphs, multiple different measures are formulated. The proposed algorithms are evaluated using the NYUv2 database. The results demonstrate that the inferred support relations are more precise than state-of-the-art. The scene graphs are compared against ground truth graphs.

arXiv:1711.06032

Natural Language Guided Visual Relationship Detection

Reasoning about the relationships between object pairs in images is a crucial task for holistic scene understanding. Most of the existing works treat this task as a pure visual classification task: each type of relationship or phrase is classified as a relation category based on the extracted visual features. However, each kind of relationships has a wide variety of object combination and each pair of objects has diverse interactions. Obtaining sufficient training samples for all possible relationship categories is difficult and expensive. In this work, we propose a natural language guided framework to tackle this problem. We propose to use a generic bi-directional recurrent neural network to predict the semantic connection between the participating objects in the relationship from the aspect of natural language. The proposed simple method achieves the state-of-the-art on the Visual Relationship Detection (VRD) and Visual Genome datasets, especially when predicting unseen relationships (e.g. recall improved from 76.42% to 89.79% on VRD zero-shot testing set).

arXiv:1711.10693

Small Drone Field Experiment: Data Collection & Processing

Following an initiative formalized in April 2016 formally known as ARL West between the U.S. Army Research Laboratory (ARL) and University of Southern California's Institute for Creative Technologies (USC ICT), a field experiment was coordinated and executed in the summer of 2016 by ARL, USC ICT, and Headwall Photonics. The purpose was to image part of the USC main campus in Los Angeles, USA, using two portable COTS (commercial off the shelf) aerial drone solutions for data acquisition, for photogrammetry (3D reconstruction from images), and fusion of hyperspectral data with the recovered set of 3D point clouds representing the target area. The research aims for determining the viability of having a machine capable of segmenting the target area into key material classes (e.g., manmade structures, live vegetation, water) for use in multiple purposes, to include providing the user with a more accurate scene understanding and enabling the unsupervised automatic sampling of meaningful material classes from the target area for adaptive semi-supervised machine learning. In the latter, a target set library may be used for automatic machine training with data of local material classes, as an example, to increase the prediction chances of machines recognizing targets. The field experiment and associated data post processing approach to correct for reflectance, geo-rectify, recover the area's dense point clouds from images, register spectral with elevation properties of scene surfaces from the independently collected datasets, and generate the desired scene segmented maps are discussed. Lessons learned from the experience are also highlighted throughout the paper.

arXiv:1712.00028

Feature discovery and visualization of robot mission data using convolutional autoencoders and Bayesian nonparametric topic models

The gap between our ability to collect interesting data and our ability to analyze these data is growing at an unprecedented rate. Recent algorithmic attempts to fill this gap have employed unsupervised tools to discover structure in data. Some of the most successful approaches have used probabilistic models to uncover latent thematic structure in discrete data. Despite the success of these models on textual data, they have not generalized as well to image data, in part because of the spatial and temporal structure that may exist in an image stream. We introduce a novel unsupervised machine learning framework that incorporates the ability of convolutional autoencoders to discover features from images that directly encode spatial information, within a Bayesian nonparametric topic model that discovers meaningful latent patterns within discrete data. By using this hybrid framework, we overcome the fundamental dependency of traditional topic models on rigidly hand-coded data representations, while simultaneously encoding spatial dependency in our topics without adding model complexity. We apply this model to the motivating application of high-level scene understanding and mission summarization for exploratory marine robots. Our experiments on a seafloor dataset collected by a marine robot show that the proposed hybrid framework outperforms current state-of-the-art approaches on the task of unsupervised seafloor terrain characterization.

arXiv:1712.01195

Why my photos look sideways or upside down? Detecting Canonical Orientation of Images using Convolutional Neural Networks

Image orientation detection requires high-level scene understanding. Humans use object recognition and contextual scene information to correctly orient images. In literature, the problem of image orientation detection is mostly confronted by using low-level vision features, while some approaches incorporate few easily detectable semantic cues to gain minor improvements. The vast amount of semantic content in images makes orientation detection challenging, and therefore there is a large semantic gap between existing methods and human behavior. Also, existing methods in literature report highly discrepant detection rates, which is mainly due to large differences in datasets and limited variety of test images used for evaluation. In this work, for the first time, we leverage the power of deep learning and adapt pre-trained convolutional neural networks using largest training dataset to-date for the image orientation detection task. An extensive evaluation of our model on different public datasets shows that it remarkably generalizes to correctly orient a large set of unconstrained images; it also significantly outperforms the state-of-the-art and achieves accuracy very close to that of humans.

arXiv:1703.08866

Multi-View Deep Learning for Consistent Semantic Mapping with RGB-D Cameras

Visual scene understanding is an important capability that enables robots to purposefully act in their environment. In this paper, we propose a novel approach to object-class segmentation from multiple RGB-D views using deep learning. We train a deep neural network to predict object-class semantics that is consistent from several view points in a semi-supervised way. At test time, the semantics predictions of our network can be fused more consistently in semantic keyframe maps than predictions of a network trained on individual views. We base our network architecture on a recent single-view deep learning approach to RGB and depth fusion for semantic object-class segmentation and enhance it with multi-scale loss minimization. We obtain the camera trajectory using RGB-D SLAM and warp the predictions of RGB-D images into ground-truth annotated frames in order to enforce multi-view consistency during training. At test time, predictions from multiple views are fused into keyframes. We propose and analyze several methods for enforcing multi-view consistency during training and testing. We evaluate the benefit of multi-view consistency training and demonstrate that pooling of deep features and fusion over multiple views outperforms single-view baselines on the NYUDv2 benchmark for semantic segmentation. Our end-to-end trained network achieves state-of-the-art performance on the NYUDv2 dataset in single-view segmentation as well as multi-view semantic fusion.

arXiv:1709.08480

J-MOD$^{2}$: Joint Monocular Obstacle Detection and Depth Estimation

In this work, we propose an end-to-end deep architecture that jointly learns to detect obstacles and estimate their depth for MAV flight applications. Most of the existing approaches either rely on Visual SLAM systems or on depth estimation models to build 3D maps and detect obstacles. However, for the task of avoiding obstacles this level of complexity is not required. Recent works have proposed multi task architectures to both perform scene understanding and depth estimation. We follow their track and propose a specific architecture to jointly estimate depth and obstacles, without the need to compute a global map, but maintaining compatibility with a global SLAM system if needed. The network architecture is devised to exploit the joint information of the obstacle detection task, that produces more reliable bounding boxes, with the depth estimation one, increasing the robustness of both to scenario changes. We call this architecture J-MOD$^{2}$. We test the effectiveness of our approach with experiments on sequences with different appearance and focal lengths and compare it to SotA multi task methods that jointly perform semantic segmentation and depth estimation. In addition, we show the integration in a full system using a set of simulated navigation experiments where a MAV explores an unknown scenario and plans safe trajectories by using our detection model.

arXiv:1712.06080공식 코드 · Papers with Code

Spatial As Deep: Spatial CNN for Traffic Scene Understanding

Convolutional neural networks (CNNs) are usually built by stacking convolutional operations layer-by-layer. Although CNN has shown strong capability to extract semantics from raw pixels, its capacity to capture spatial relationships of pixels across rows and columns of an image is not fully explored. These relationships are important to learn semantic objects with strong shape priors but weak appearance coherences, such as traffic lanes, which are often occluded or not even painted on the road surface as shown in Fig. 1 (a). In this paper, we propose Spatial CNN (SCNN), which generalizes traditional deep layer-by-layer convolutions to slice-byslice convolutions within feature maps, thus enabling message passings between pixels across rows and columns in a layer. Such SCNN is particular suitable for long continuous shape structure or large objects, with strong spatial relationship but less appearance clues, such as traffic lanes, poles, and wall. We apply SCNN on a newly released very challenging traffic lane detection dataset and Cityscapse dataset. The results show that SCNN could learn the spatial relationship for structure output and significantly improves the performance. We show that SCNN outperforms the recurrent neural network (RNN) based ReNet and MRF+CNN (MRFNet) in the lane detection dataset by 8.7% and 4.6% respectively. Moreover, our SCNN won the 1st place on the TuSimple Benchmark Lane Detection Challenge, with an accuracy of 96.53%.

arXiv:1801.01281

Object segmentation in depth maps with one user click and a synthetically trained fully convolutional network

With more and more household objects built on planned obsolescence and consumed by a fast-growing population, hazardous waste recycling has become a critical challenge. Given the large variability of household waste, current recycling platforms mostly rely on human operators to analyze the scene, typically composed of many object instances piled up in bulk. Helping them by robotizing the unitary extraction is a key challenge to speed up this tedious process. Whereas supervised deep learning has proven very efficient for such object-level scene understanding, e.g., generic object detection and segmentation in everyday scenes, it however requires large sets of per-pixel labeled images, that are hardly available for numerous application contexts, including industrial robotics. We thus propose a step towards a practical interactive application for generating an object-oriented robotic grasp, requiring as inputs only one depth map of the scene and one user click on the next object to extract. More precisely, we address in this paper the middle issue of object seg-mentation in top views of piles of bulk objects given a pixel location, namely seed, provided interactively by a human operator. We propose a twofold framework for generating edge-driven instance segments. First, we repurpose a state-of-the-art fully convolutional object contour detector for seed-based instance segmentation by introducing the notion of edge-mask duality with a novel patch-free and contour-oriented loss function. Second, we train one model using only synthetic scenes, instead of manually labeled training data. Our experimental results show that considering edge-mask duality for training an encoder-decoder network, as we suggest, outperforms a state-of-the-art patch-based network in the present application context.

arXiv:1801.01235

Depth Not Needed - An Evaluation of RGB-D Feature Encodings for Off-Road Scene Understanding by Convolutional Neural Network

Scene understanding for autonomous vehicles is a challenging computer vision task, with recent advances in convolutional neural networks (CNNs) achieving results that notably surpass prior traditional feature driven approaches. However, limited work investigates the application of such methods either within the highly unstructured off-road environment or to RGBD input data. In this work, we take an existing CNN architecture designed to perform semantic segmentation of RGB images of urban road scenes, then adapt and retrain it to perform the same task with multichannel RGBD images obtained under a range of challenging off-road conditions. We compare two different stereo matching algorithms and five different methods of encoding depth information, including disparity, local normal orientation and HHA (horizontal disparity, height above ground plane, angle with gravity), to create a total of ten experimental variations of our dataset, each of which is used to train and test a CNN so that classification performance can be evaluated against a CNN trained using standard RGB input.

arXiv:1801.03127

Recognizing Material Properties from Images

Humans rely on properties of the materials that make up objects to guide our interactions with them. Grasping smooth materials, for example, requires care, and softness is an ideal property for fabric used in bedding. Even when these properties are not visual (e.g. softness is a physical property), we may still infer their presence visually. We refer to such material properties as visual material attributes. Recognizing these attributes in images can contribute valuable information for general scene understanding and material recognition. Unlike well-known object and scene attributes, visual material attributes are local properties with no fixed shape or spatial extent. We show that given a set of images annotated with known material attributes, we may accurately recognize the attributes from small local image patches. Obtaining such annotations in a consistent fashion at scale, however, is challenging. To address this, we introduce a method that allows us to probe the human visual perception of materials by asking simple yes/no questions comparing pairs of image patches. This provides sufficient weak supervision to build a set of attributes and associated classifiers that, while unnamed, serve the same function as the named attributes we use to describe materials. Doing so allows us to recognize visual material attributes without resorting to exhaustive manual annotation of a fixed set of named attributes. Furthermore, we show that this method may be integrated in the end-to-end learning of a material classification CNN to simultaneously recognize materials and discover their visual attributes. Our experimental results show that visual material attributes, whether named or automatically discovered, provide a useful intermediate representation for known material categories themselves as well as a basis for transfer learning when recognizing previously-unseen categories.

arXiv:1605.09332

Parametric Exponential Linear Unit for Deep Convolutional Neural Networks

Object recognition is an important task for improving the ability of visual systems to perform complex scene understanding. Recently, the Exponential Linear Unit (ELU) has been proposed as a key component for managing bias shift in Convolutional Neural Networks (CNNs), but defines a parameter that must be set by hand. In this paper, we propose learning a parameterization of ELU in order to learn the proper activation shape at each layer in the CNNs. Our results on the MNIST, CIFAR-10/100 and ImageNet datasets using the NiN, Overfeat, All-CNN and ResNet networks indicate that our proposed Parametric ELU (PELU) has better performances than the non-parametric ELU. We have observed as much as a 7.28% relative error improvement on ImageNet with the NiN network, with only 0.0003% parameter increase. Our visual examination of the non-linear behaviors adopted by Vgg using PELU shows that the network took advantage of the added flexibility by learning different activations at different layers.

arXiv:1709.09490

Hierarchical Scene Parsing by Weakly Supervised Learning with Image Descriptions

This paper investigates a fundamental problem of scene understanding: how to parse a scene image into a structured configuration (i.e., a semantic object hierarchy with object interaction relations). We propose a deep architecture consisting of two networks: i) a convolutional neural network (CNN) extracting the image representation for pixel-wise object labeling and ii) a recursive neural network (RsNN) discovering the hierarchical object structure and the inter-object relations. Rather than relying on elaborative annotations (e.g., manually labeled semantic maps and relations), we train our deep model in a weakly-supervised learning manner by leveraging the descriptive sentences of the training images. Specifically, we decompose each sentence into a semantic tree consisting of nouns and verb phrases, and apply these tree structures to discover the configurations of the training images. Once these scene configurations are determined, then the parameters of both the CNN and RsNN are updated accordingly by back propagation. The entire model training is accomplished through an Expectation-Maximization method. Extensive experiments show that our model is capable of producing meaningful scene configurations and achieving more favorable scene labeling results on two benchmarks (i.e., PASCAL VOC 2012 and SYSU-Scenes) compared with other state-of-the-art weakly-supervised deep learning methods. In particular, SYSU-Scenes contains more than 5000 scene images with their semantic sentence descriptions, which is created by us for advancing research on scene parsing.

arXiv:1802.06117

Scenarios: A New Representation for Complex Scene Understanding

The ability for computational agents to reason about the high-level content of real world scene images is important for many applications. Existing attempts at addressing the problem of complex scene understanding lack representational power, efficiency, and the ability to create robust meta-knowledge about scenes. In this paper, we introduce scenarios as a new way of representing scenes. The scenario is a simple, low-dimensional, data-driven representation consisting of sets of frequently co-occurring objects and is useful for a wide range of scene understanding tasks. We learn scenarios from data using a novel matrix factorization method which we integrate into a new neural network architecture, the ScenarioNet. Using ScenarioNet, we can recover semantic information about real world scene images at three levels of granularity: 1) scene categories, 2) scenarios, and 3) objects. Training a single ScenarioNet model enables us to perform scene classification, scenario recognition, multi-object recognition, content-based scene image retrieval, and content-based image comparison. In addition to solving many tasks in a single, unified framework, ScenarioNet is more computationally efficient than other CNNs because it requires significantly fewer parameters while achieving similar performance on benchmark tasks and is more interpretable because it produces explanations when making decisions. We validate the utility of scenarios and ScenarioNet on a diverse set of scene understanding tasks on several benchmark datasets.

arXiv:1802.06459

Structured Label Inference for Visual Understanding

Visual data such as images and videos contain a rich source of structured semantic labels as well as a wide range of interacting components. Visual content could be assigned with fine-grained labels describing major components, coarse-grained labels depicting high level abstractions, or a set of labels revealing attributes. Such categorization over different, interacting layers of labels evinces the potential for a graph-based encoding of label information. In this paper, we exploit this rich structure for performing graph-based inference in label space for a number of tasks: multi-label image and video classification and action detection in untrimmed videos. We consider the use of the Bidirectional Inference Neural Network (BINN) and Structured Inference Neural Network (SINN) for performing graph-based inference in label space and propose a Long Short-Term Memory (LSTM) based extension for exploiting activity progression on untrimmed videos. The methods were evaluated on (i) the Animal with Attributes (AwA), Scene Understanding (SUN) and NUS-WIDE datasets for multi-label image classification, (ii) the first two releases of the YouTube-8M large scale dataset for multi-label video classification, and (iii) the THUMOS'14 and MultiTHUMOS video datasets for action detection. Our results demonstrate the effectiveness of structured label inference in these challenging tasks, achieving significant improvements against baselines.