PyTorchKR Papers

최신 논문

총 1,581편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2203.09375

Neural Part Priors: Learning to Optimize Part-Based Object Completion in RGB-D Scans

3D object recognition has seen significant advances in recent years, showing impressive performance on real-world 3D scan benchmarks, but lacking in object part reasoning, which is fundamental to higher-level scene understanding such as inter-object similarities or object functionality. Thus, we propose to leverage large-scale synthetic datasets of 3D shapes annotated with part information to learn Neural Part Priors (NPPs), optimizable spaces characterizing geometric part priors. Crucially, we can optimize over the learned part priors in order to fit to real-world scanned 3D scenes at test time, enabling robust part decomposition of the real objects in these scenes that also estimates the complete geometry of the object while fitting accurately to the observed real geometry. Moreover, this enables global optimization over geometrically similar detected objects in a scene, which often share strong geometric commonalities, enabling scene-consistent part decompositions. Experiments on the ScanNet dataset demonstrate that NPPs significantly outperforms state of the art in part decomposition and object completion in real-world scenes.

arXiv:2203.09138공식 코드 · Papers with Code

MuKEA: Multimodal Knowledge Extraction and Accumulation for Knowledge-based Visual Question Answering

Knowledge-based visual question answering requires the ability of associating external knowledge for open-ended cross-modal scene understanding. One limitation of existing solutions is that they capture relevant knowledge from text-only knowledge bases, which merely contain facts expressed by first-order predicates or language descriptions while lacking complex but indispensable multimodal knowledge for visual understanding. How to construct vision-relevant and explainable multimodal knowledge for the VQA scenario has been less studied. In this paper, we propose MuKEA to represent multimodal knowledge by an explicit triplet to correlate visual objects and fact answers with implicit relations. To bridge the heterogeneous gap, we propose three objective losses to learn the triplet representations from complementary views: embedding structure, topological relation and semantic space. By adopting a pre-training and fine-tuning learning strategy, both basic and domain-specific multimodal knowledge are progressively accumulated for answer prediction. We outperform the state-of-the-art by 3.35% and 6.08% respectively on two challenging knowledge-required datasets: OK-VQA and KRVQA. Experimental results prove the complementary benefits of the multimodal knowledge with existing knowledge bases and the advantages of our end-to-end framework over the existing pipeline methods. The code is available at https://github.com/AndersonStra/MuKEA.

arXiv:2203.09088

Deep Point Cloud Simplification for High-quality Surface Reconstruction

The growing size of point clouds enlarges consumptions of storage, transmission, and computation of 3D scenes. Raw data is redundant, noisy, and non-uniform. Therefore, simplifying point clouds for achieving compact, clean, and uniform points is becoming increasingly important for 3D vision and graphics tasks. Previous learning based methods aim to generate fewer points for scene understanding, regardless of the quality of surface reconstruction, leading to results with low reconstruction accuracy and bad point distribution. In this paper, we propose a novel point cloud simplification network (PCS-Net) dedicated to high-quality surface mesh reconstruction while maintaining geometric fidelity. We first learn a sampling matrix in a feature-aware simplification module to reduce the number of points. Then we propose a novel double-scale resampling module to refine the positions of the sampled points, to achieve a uniform distribution. To further retain important shape features, an adaptive sampling strategy with a novel saliency loss is designed. With our PCS-Net, the input non-uniform and noisy point cloud can be simplified in a feature-aware manner, i.e., points near salient features are consolidated but still with uniform distribution locally. Experiments demonstrate the effectiveness of our method and show that we outperform previous simplification or reconstruction-oriented upsampling methods.

arXiv:2203.08332공식 코드 · Papers with Code

WeakM3D: Towards Weakly Supervised Monocular 3D Object Detection

Monocular 3D object detection is one of the most challenging tasks in 3D scene understanding. Due to the ill-posed nature of monocular imagery, existing monocular 3D detection methods highly rely on training with the manually annotated 3D box labels on the LiDAR point clouds. This annotation process is very laborious and expensive. To dispense with the reliance on 3D box labels, in this paper we explore the weakly supervised monocular 3D detection. Specifically, we first detect 2D boxes on the image. Then, we adopt the generated 2D boxes to select corresponding RoI LiDAR points as the weak supervision. Eventually, we adopt a network to predict 3D boxes which can tightly align with associated RoI LiDAR points. This network is learned by minimizing our newly-proposed 3D alignment loss between the 3D box estimates and the corresponding RoI LiDAR points. We will illustrate the potential challenges of the above learning problem and resolve these challenges by introducing several effective designs into our method. Codes will be available at https://github.com/SPengLiang/WeakM3D.

arXiv:2203.08038공식 코드 · Papers with Code

Deep learning for radar data exploitation of autonomous vehicle

Autonomous driving requires a detailed understanding of complex driving scenes. The redundancy and complementarity of the vehicle's sensors provide an accurate and robust comprehension of the environment, thereby increasing the level of performance and safety. This thesis focuses the on automotive RADAR, which is a low-cost active sensor measuring properties of surrounding objects, including their relative speed, and has the key advantage of not being impacted by adverse weather conditions. With the rapid progress of deep learning and the availability of public driving datasets, the perception ability of vision-based driving systems has considerably improved. The RADAR sensor is seldom used for scene understanding due to its poor angular resolution, the size, noise, and complexity of RADAR raw data as well as the lack of available datasets. This thesis proposes an extensive study of RADAR scene understanding, from the construction of an annotated dataset to the conception of adapted deep learning architectures. First, this thesis details approaches to tackle the current lack of data. A simple simulation as well as generative methods for creating annotated data will be presented. It will also describe the CARRADA dataset, composed of synchronised camera and RADAR data with a semi-automatic annotation method. This thesis then present a proposed set of deep learning architectures with their associated loss functions for RADAR semantic segmentation. It also introduces a method to open up research into the fusion of LiDAR and RADAR sensors for scene understanding. Finally, this thesis exposes a collaborative contribution, the RADIal dataset with synchronised High-Definition (HD) RADAR, LiDAR and camera. A deep learning architecture is also proposed to estimate the RADAR signal processing pipeline while performing multitask learning for object detection and free driving space segmentation.

arXiv:2203.07997공식 코드 · Papers with Code

InvPT: Inverted Pyramid Multi-task Transformer for Dense Scene Understanding

Multi-task dense scene understanding is a thriving research domain that requires simultaneous perception and reasoning on a series of correlated tasks with pixel-wise prediction. Most existing works encounter a severe limitation of modeling in the locality due to heavy utilization of convolution operations, while learning interactions and inference in a global spatial-position and multi-task context is critical for this problem. In this paper, we propose a novel end-to-end Inverted Pyramid multi-task Transformer (InvPT) to perform simultaneous modeling of spatial positions and multiple tasks in a unified framework. To the best of our knowledge, this is the first work that explores designing a transformer structure for multi-task dense prediction for scene understanding. Besides, it is widely demonstrated that a higher spatial resolution is remarkably beneficial for dense predictions, while it is very challenging for existing transformers to go deeper with higher resolutions due to huge complexity to large spatial size. InvPT presents an efficient UP-Transformer block to learn multi-task feature interaction at gradually increased resolutions, which also incorporates effective self-attention message passing and multi-scale feature aggregation to produce task-specific prediction at a high resolution. Our method achieves superior multi-task performance on NYUD-v2 and PASCAL-Context datasets respectively, and significantly outperforms previous state-of-the-arts. The code is available at https://github.com/prismformore/InvPT

arXiv:2203.07162

RAUM-VO: Rotational Adjusted Unsupervised Monocular Visual Odometry

Unsupervised learning for monocular camera motion and 3D scene understanding has gained popularity over traditional methods, relying on epipolar geometry or non-linear optimization. Notably, deep learning can overcome many issues of monocular vision, such as perceptual aliasing, low-textured areas, scale-drift, and degenerate motions. Also, concerning supervised learning, we can fully leverage video streams data without the need for depth or motion labels. However, in this work, we note that rotational motion can limit the accuracy of the unsupervised pose networks more than the translational component. Therefore, we present RAUM-VO, an approach based on a model-free epipolar constraint for frame-to-frame motion estimation (F2F) to adjust the rotation during training and online inference. To this end, we match 2D keypoints between consecutive frames using pre-trained deep networks, Superpoint and Superglue, while training a network for depth and pose estimation using an unsupervised training protocol. Then, we adjust the predicted rotation with the motion estimated by F2F using the 2D matches and initializing the solver with the pose network prediction. Ultimately, RAUM-VO shows a considerable accuracy improvement compared to other unsupervised pose networks on the KITTI dataset while reducing the complexity of other hybrid or traditional approaches and achieving comparable state-of-the-art results.

arXiv:2203.04838공식 코드 · Papers with Code

CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with Transformers

Scene understanding based on image segmentation is a crucial component of autonomous vehicles. Pixel-wise semantic segmentation of RGB images can be advanced by exploiting complementary features from the supplementary modality (X-modality). However, covering a wide variety of sensors with a modality-agnostic model remains an unresolved problem due to variations in sensor characteristics among different modalities. Unlike previous modality-specific methods, in this work, we propose a unified fusion framework, CMX, for RGB-X semantic segmentation. To generalize well across different modalities, that often include supplements as well as uncertainties, a unified cross-modal interaction is crucial for modality fusion. Specifically, we design a Cross-Modal Feature Rectification Module (CM-FRM) to calibrate bi-modal features by leveraging the features from one modality to rectify the features of the other modality. With rectified feature pairs, we deploy a Feature Fusion Module (FFM) to perform sufficient exchange of long-range contexts before mixing. To verify CMX, for the first time, we unify five modalities complementary to RGB, i.e., depth, thermal, polarization, event, and LiDAR. Extensive experiments show that CMX generalizes well to diverse multi-modal fusion, achieving state-of-the-art performances on five RGB-Depth benchmarks, as well as RGB-Thermal, RGB-Polarization, and RGB-LiDAR datasets. Besides, to investigate the generalizability to dense-sparse data fusion, we establish an RGB-Event semantic segmentation benchmark based on the EventScape dataset, on which CMX sets the new state-of-the-art. The source code of CMX is publicly available at https://github.com/huaaaliu/RGBX_Semantic_Segmentation.

arXiv:2203.02946

On Steering Multi-Annotations per Sample for Multi-Task Learning

The study of multi-task learning has drawn great attention from the community. Despite the remarkable progress, the challenge of optimally learning different tasks simultaneously remains to be explored. Previous works attempt to modify the gradients from different tasks. Yet these methods give a subjective assumption of the relationship between tasks, and the modified gradient may be less accurate. In this paper, we introduce Stochastic Task Allocation~(STA), a mechanism that addresses this issue by a task allocation approach, in which each sample is randomly allocated a subset of tasks. For further progress, we propose Interleaved Stochastic Task Allocation~(ISTA) to iteratively allocate all tasks to each example during several consecutive iterations. We evaluate STA and ISTA on various datasets and applications: NYUv2, Cityscapes, and COCO for scene understanding and instance segmentation. Our experiments show both STA and ISTA outperform current state-of-the-art methods. The code will be available.

arXiv:2203.01994

Fast Neural Architecture Search for Lightweight Dense Prediction Networks

We present LDP, a lightweight dense prediction neural architecture search (NAS) framework. Starting from a pre-defined generic backbone, LDP applies the novel Assisted Tabu Search for efficient architecture exploration. LDP is fast and suitable for various dense estimation problems, unlike previous NAS methods that are either computational demanding or deployed only for a single subtask. The performance of LPD is evaluated on monocular depth estimation, semantic segmentation, and image super-resolution tasks on diverse datasets, including NYU-Depth-v2, KITTI, Cityscapes, COCO-stuff, DIV2K, Set5, Set14, BSD100, Urban100. Experiments show that the proposed framework yields consistent improvements on all tested dense prediction tasks, while being $5\%-315\%$ more compact in terms of the number of model parameters than prior arts.

arXiv:2203.01452공식 코드 · Papers with Code

Bending Reality: Distortion-aware Transformers for Adapting to Panoramic Semantic Segmentation

Panoramic images with their 360-degree directional view encompass exhaustive information about the surrounding space, providing a rich foundation for scene understanding. To unfold this potential in the form of robust panoramic segmentation models, large quantities of expensive, pixel-wise annotations are crucial for success. Such annotations are available, but predominantly for narrow-angle, pinhole-camera images which, off the shelf, serve as sub-optimal resources for training panoramic models. Distortions and the distinct image-feature distribution in 360-degree panoramas impede the transfer from the annotation-rich pinhole domain and therefore come with a big dent in performance. To get around this domain difference and bring together semantic annotations from pinhole- and 360-degree surround-visuals, we propose to learn object deformations and panoramic image distortions in the Deformable Patch Embedding (DPE) and Deformable MLP (DMLP) components which blend into our Transformer for PAnoramic Semantic Segmentation (Trans4PASS) model. Finally, we tie together shared semantics in pinhole- and panoramic feature embeddings by generating multi-scale prototype features and aligning them in our Mutual Prototypical Adaptation (MPA) for unsupervised domain adaptation. On the indoor Stanford2D3D dataset, our Trans4PASS with MPA maintains comparable performance to fully-supervised state-of-the-arts, cutting the need for over 1,400 labeled panoramas. On the outdoor DensePASS dataset, we break state-of-the-art by 14.39% mIoU and set the new bar at 56.38%. Code will be made publicly available at https://github.com/jamycheung/Trans4PASS.

arXiv:2203.00869

Hybrid Optimized Deep Convolution Neural Network based Learning Model for Object Detection

Object identification is one of the most fundamental and difficult issues in computer vision. It aims to discover object instances in real pictures from a huge number of established categories. In recent years, deep learning-based object detection techniques that developed from computer vision have grabbed the public's interest. Object recognition methods based on deep learning frameworks have quickly become a popular way to interpret moving images acquired by various sensors. Due to its vast variety of applications for various computer vision tasks such as activity or event detection, content-based image retrieval, and scene understanding, academics have spent decades attempting to solve this problem. With this goal in mind, a unique deep learning classification technique is used to create an autonomous object detecting system. The noise destruction and normalising operations, which are carried out using gaussian filter and contrast normalisation techniques, respectively, are the first steps in the study activity. The pre-processed picture is next subjected to entropy-based segmentation algorithms, which separate the image's significant areas in order to distinguish between distinct occurrences. The classification challenge is completed by the suggested Hybrid Optimized Dense Convolutional Neural Network (HODCNN). The major goal of this framework is to aid in the precise recognition of distinct items from the gathered input frames. The suggested system's performance is assessed by comparing it to existing machine learning and deep learning methodologies. The experimental findings reveal that the suggested framework has a detection accuracy of 0.9864, which is greater than current techniques. As a result, the suggested object detection model outperforms other current methods.

arXiv:2202.13393공식 코드 · Papers with Code

TransKD: Transformer Knowledge Distillation for Efficient Semantic Segmentation

Semantic segmentation benchmarks in the realm of autonomous driving are dominated by large pre-trained transformers, yet their widespread adoption is impeded by substantial computational costs and prolonged training durations. To lift this constraint, we look at efficient semantic segmentation from a perspective of comprehensive knowledge distillation and aim to bridge the gap between multi-source knowledge extractions and transformer-specific patch embeddings. We put forward the Transformer-based Knowledge Distillation (TransKD) framework which learns compact student transformers by distilling both feature maps and patch embeddings of large teacher transformers, bypassing the long pre-training process and reducing the FLOPs by >85.0%. Specifically, we propose two fundamental modules to realize feature map distillation and patch embedding distillation, respectively: (1) Cross Selective Fusion (CSF) enables knowledge transfer between cross-stage features via channel attention and feature map distillation within hierarchical transformers; (2) Patch Embedding Alignment (PEA) performs dimensional transformation within the patchifying process to facilitate the patch embedding distillation. Furthermore, we introduce two optimization modules to enhance the patch embedding distillation from different perspectives: (1) Global-Local Context Mixer (GL-Mixer) extracts both global and local information of a representative embedding; (2) Embedding Assistant (EA) acts as an embedding method to seamlessly bridge teacher and student models with the teacher's number of channels. Experiments on Cityscapes, ACDC, NYUv2, and Pascal VOC2012 datasets show that TransKD outperforms state-of-the-art distillation frameworks and rivals the time-consuming pre-training method. The source code is publicly available at https://github.com/RuipingL/TransKD.

arXiv:2202.13094공식 코드 · Papers with Code

RIConv++: Effective Rotation Invariant Convolutions for 3D Point Clouds Deep Learning

3D point clouds deep learning is a promising field of research that allows a neural network to learn features of point clouds directly, making it a robust tool for solving 3D scene understanding tasks. While recent works show that point cloud convolutions can be invariant to translation and point permutation, investigations of the rotation invariance property for point cloud convolution has been so far scarce. Some existing methods perform point cloud convolutions with rotation-invariant features, existing methods generally do not perform as well as translation-invariant only counterpart. In this work, we argue that a key reason is that compared to point coordinates, rotation-invariant features consumed by point cloud convolution are not as distinctive. To address this problem, we propose a simple yet effective convolution operator that enhances feature distinction by designing powerful rotation invariant features from the local regions. We consider the relationship between the point of interest and its neighbors as well as the internal relationship of the neighbors to largely improve the feature descriptiveness. Our network architecture can capture both local and global context by simply tuning the neighborhood size in each convolution layer. We conduct several experiments on synthetic and real-world point cloud classifications, part segmentation, and shape retrieval to evaluate our method, which achieves the state-of-the-art accuracy under challenging rotations.

arXiv:2202.12361공식 코드 · Papers with Code

RescueNet: A High Resolution UAV Semantic Segmentation Benchmark Dataset for Natural Disaster Damage Assessment

Recent advancements in computer vision and deep learning techniques have facilitated notable progress in scene understanding, thereby assisting rescue teams in achieving precise damage assessment. In this paper, we present RescueNet, a meticulously curated high-resolution post-disaster dataset that includes detailed classification and semantic segmentation annotations. This dataset aims to facilitate comprehensive scene understanding in the aftermath of natural disasters. RescueNet comprises post-disaster images collected after Hurricane Michael, obtained using Unmanned Aerial Vehicles (UAVs) from multiple impacted regions. The uniqueness of RescueNet lies in its provision of high-resolution post-disaster imagery, accompanied by comprehensive annotations for each image. Unlike existing datasets that offer annotations limited to specific scene elements such as buildings, RescueNet provides pixel-level annotations for all classes, including buildings, roads, pools, trees, and more. Furthermore, we evaluate the utility of the dataset by implementing state-of-the-art segmentation models on RescueNet, demonstrating its value in enhancing existing methodologies for natural disaster damage assessment.

arXiv:2202.11094공식 코드 · Papers with Code

GroupViT: Semantic Segmentation Emerges from Text Supervision

Grouping and recognition are important components of visual scene understanding, e.g., for object detection and semantic segmentation. With end-to-end deep learning systems, grouping of image regions usually happens implicitly via top-down supervision from pixel-level recognition labels. Instead, in this paper, we propose to bring back the grouping mechanism into deep networks, which allows semantic segments to emerge automatically with only text supervision. We propose a hierarchical Grouping Vision Transformer (GroupViT), which goes beyond the regular grid structure representation and learns to group image regions into progressively larger arbitrary-shaped segments. We train GroupViT jointly with a text encoder on a large-scale image-text dataset via contrastive losses. With only text supervision and without any pixel-level annotations, GroupViT learns to group together semantic regions and successfully transfers to the task of semantic segmentation in a zero-shot manner, i.e., without any further fine-tuning. It achieves a zero-shot accuracy of 52.3% mIoU on the PASCAL VOC 2012 and 22.4% mIoU on PASCAL Context datasets, and performs competitively to state-of-the-art transfer-learning methods requiring greater levels of supervision. We open-source our code at https://github.com/NVlabs/GroupViT .

arXiv:2202.11092공식 코드 · Papers with Code

ReorientBot: Learning Object Reorientation for Specific-Posed Placement

Robots need the capability of placing objects in arbitrary, specific poses to rearrange the world and achieve various valuable tasks. Object reorientation plays a crucial role in this as objects may not initially be oriented such that the robot can grasp and then immediately place them in a specific goal pose. In this work, we present a vision-based manipulation system, ReorientBot, which consists of 1) visual scene understanding with pose estimation and volumetric reconstruction using an onboard RGB-D camera; 2) learned waypoint selection for successful and efficient motion generation for reorientation; 3) traditional motion planning to generate a collision-free trajectory from the selected waypoints. We evaluate our method using the YCB objects in both simulation and the real world, achieving 93% overall success, 81% improvement in success rate, and 22% improvement in execution time compared to a heuristic approach. We demonstrate extended multi-object rearrangement showing the general capability of the system.

arXiv:2202.10650

Movies2Scenes: Using Movie Metadata to Learn Scene Representation

Understanding scenes in movies is crucial for a variety of applications such as video moderation, search, and recommendation. However, labeling individual scenes is a time-consuming process. In contrast, movie level metadata (e.g., genre, synopsis, etc.) regularly gets produced as part of the film production process, and is therefore significantly more commonly available. In this work, we propose a novel contrastive learning approach that uses movie metadata to learn a general-purpose scene representation. Specifically, we use movie metadata to define a measure of movie similarity, and use it during contrastive learning to limit our search for positive scene-pairs to only the movies that are considered similar to each other. Our learned scene representation consistently outperforms existing state-of-the-art methods on a diverse set of tasks evaluated using multiple benchmark datasets. Notably, our learned representation offers an average improvement of 7.9% on the seven classification tasks and 9.7% improvement on the two regression tasks in LVU dataset. Furthermore, using a newly collected movie dataset, we present comparative results of our scene representation on a set of video moderation tasks to demonstrate its generalizability on previously less explored tasks.

arXiv:2202.09721공식 코드 · Papers with Code

3DRM:Pair-wise relation module for 3D object detection

Context has proven to be one of the most important factors in object layout reasoning for 3D scene understanding. Existing deep contextual models either learn holistic features for context encoding or rely on pre-defined scene templates for context modeling. We argue that scene understanding benefits from object relation reasoning, which is capable of mitigating the ambiguity of 3D object detections and thus helps locate and classify the 3D objects more accurately and robustly. To achieve this, we propose a novel 3D relation module (3DRM) which reasons about object relations at pair-wise levels. The 3DRM predicts the semantic and spatial relationships between objects and extracts the object-wise relation features. We demonstrate the effects of 3DRM by plugging it into proposal-based and voting-based 3D object detection pipelines, respectively. Extensive evaluations show the effectiveness and generalization of 3DRM on 3D object detection. Our source code is available at https://github.com/lanlan96/3DRM.

arXiv:2202.06851공식 코드 · Papers with Code

HAKE: A Knowledge Engine Foundation for Human Activity Understanding

Human activity understanding is of widespread interest in artificial intelligence and spans diverse applications like health care and behavior analysis. Although there have been advances in deep learning, it remains challenging. The object recognition-like solutions usually try to map pixels to semantics directly, but activity patterns are much different from object patterns, thus hindering success. In this work, we propose a novel paradigm to reformulate this task in two stages: first mapping pixels to an intermediate space spanned by atomic activity primitives, then programming detected primitives with interpretable logic rules to infer semantics. To afford a representative primitive space, we build a knowledge base including 26+ M primitive labels and logic rules from human priors or automatic discovering. Our framework, the Human Activity Knowledge Engine (HAKE), exhibits superior generalization ability and performance upon canonical methods on challenging benchmarks. Code and data are available at http://hake-mvig.cn/.