Generating coherent and useful image/video scenes from a free-form textual description is technically a very difficult problem to handle. Textual description of the same scene can vary greatly from person to person, or sometimes even for the same person from time to time. As the choice of words and syntax vary while preparing a textual description, it is challenging for the system to reliably produce a consistently desirable output from different forms of language input. The prior works of scene generation have been mostly confined to rigorous sentence structures of text input which restrict the freedom of users to write description. In our work, we study a new pipeline that aims to generate static as well as animated 3D scenes from different types of free-form textual scene description without any major restriction. In particular, to keep our study practical and tractable, we focus on a small subspace of all possible 3D scenes, containing various combinations of cube, cylinder and sphere. We design a two-stage pipeline. In the first stage, we encode the free-form text using an encoder-decoder neural architecture. In the second stage, we generate a 3D scene based on the generated encoding. Our neural architecture exploits state-of-the-art language model as encoder to leverage rich contextual encoding and a new multi-head decoder to predict multiple features of an object in the scene simultaneously. For our experiments, we generate a large synthetic data-set which contains 13,00,000 and 14,00,000 samples of unique static and animated scene descriptions, respectively. We achieve 98.427% accuracy on test data set in detecting the 3D objects features successfully. Our work shows a proof of concept of one approach towards solving the problem, and we believe with enough training data, the same pipeline can be expanded to handle even broader set of 3D scene generation problems.
Advances in the state of the art for 3d human sensing are currently limited by the lack of visual datasets with 3d ground truth, including multiple people, in motion, operating in real-world environments, with complex illumination or occlusion, and potentially observed by a moving camera. Sophisticated scene understanding would require estimating human pose and shape as well as gestures, towards representations that ultimately combine useful metric and behavioral signals with free-viewpoint photo-realistic visualisation capabilities. To sustain progress, we build a large-scale photo-realistic …
In this paper, we demonstrate a fully automatic method for converting a still image into a realistic animated looping video. We target scenes with continuous fluid motion, such as flowing water and billowing smoke. Our method relies on the observation that this type of natural motion can be convincingly reproduced from a static Eulerian motion description, i.e. a single, temporally constant flow field that defines the immediate motion of a particle at a given 2D location. We use an image-to-image translation network to encode motion priors of natural scenes collected from online videos, so tha…
We propose a learning-based framework for disentangling outdoor scenes into temporally-varying illumination and permanent scene factors. Inspired by the classic intrinsic image decomposition, our learning signal builds upon two insights: 1) combining the disentangled factors should reconstruct the original image, and 2) the permanent factors should stay constant across multiple temporal samples of the same scene. To facilitate training, we assemble a city-scale dataset of outdoor timelapse imagery from Google Street View, where the same locations are captured repeatedly through time. This data…
KTH Royal Institute of Technology（瑞典皇家理工学院）；Mitsubishi Electric Research Labs（三菱电机研究实验室）
We present a technique for simultaneous 3D reconstruction of static regions and rigidly moving objects in a scene. An RGB-D frame is represented as a collection of features, which are points and planes. We classify the features into static and dynamic regions and grow separate maps, static and object maps, for each of them. To robustly classify the features in each frame, we fuse multiple RANSAC-based registration results obtained by registering different groups of the features to different maps, including (1) all the features to the static map, (2) all the features to each object map, and (3)…
We hypothesize that an agent that can look around in static scenes can learn rich visual representations applicable to 3D object tracking in complex dynamic scenes. We are motivated in this pursuit by the fact that the physical world itself is mostly static, and multiview correspondence labels are relatively cheap to collect in static scenes, e.g., by triangulation. We propose to leverage multiview data of \textit{static points} in arbitrary scenes (static or dynamic), to learn a neural 3D mapping module which produces features that are correspondable across time. The neural 3D mapper consumes…
Noise2Image: Noise-Enabled Static Scene Recovery for Event Cameras
Department of Bioengineering, University of California, Berkeley（生物工程系，加州大学伯克利分校）；Department of Electrical Engineering and Computer Sciences, University of California, Berkeley（电气工程与计算机科学系，加州大学伯克利分校）；Department of Vision Science, University of California, Berkeley（视觉科学系，加州大学伯克利分校）
Efficient 3D Reconstruction, Streaming and Visualization of Static and Dynamic Scene Parts for Multi-client Live-telepresence in Large-scale Environments
University of Bonn（波恩大学）；Delft University of Technology（代尔夫特理工大学）
Representing and synthesizing novel views in real-world dynamic scenes from casual monocular videos is a long-standing problem. Existing solutions typically approach dynamic scenes by applying geometry techniques or utilizing temporal information between several adjacent frames without considering the underlying background distribution in the entire scene or the transmittance over the ray dimension, limiting their performance on static and occlusion areas. Our approach $\textbf{D}$istribution-$\textbf{D}$riven neural radiance fields offers high-quality view synthesis and a 3D solution to $\tex…
GA-GS: Generation-Assisted Gaussian Splatting for Static Scene Reconstruction
School of Computer Science and Technology, University of Science and Technology of China（中国科学技术大学计算机科学与技术学院）；School of Artificial Intelligence and Data Science, University of Science and Technology of China（中国科学技术大学人工智能与数据科学学院）；The Chinese University of Hong Kong（香港中文大学）；Institute of Advanced Technology, University of Science and Technology of China（中国科学技术大学先进技术研究院）；Institute of Artificial Intelligence, Hefei Comprehensive National Science Center（合肥综合性国家科学中心人工智能研究院）；School of information Science and Technology, University of Science and Technology of China（中国科学技术大学信息科学技术学院）
There is a general expectation that robots should operate in environments that consist of static and dynamic entities including people, furniture and automobiles. These dynamic environments pose challenges to visual simultaneous localization and mapping (SLAM) algorithms by introducing errors into the front-end. Light fields provide one possible method for addressing such problems by capturing a more complete visual information of a scene. In contrast to a single ray from a perspective camera, Light Fields capture a bundle of light rays emerging from a single point in space, allowing us to see…
Ground based long-range passive imaging systems often suffer from degraded image quality due to a turbulent atmosphere. While methods exist for removing such turbulent distortions, many are limited to static sequences which cannot be extended to dynamic scenes. In addition, the physics of the turbulence is often not integrated into the image reconstruction algorithms, making the physics foundations of the methods weak. In this paper, we present a unified method for atmospheric turbulence mitigation in both static and dynamic sequences. We are able to achieve better results compared to existing…
We present a novel neural radiance model that is trainable in a self-supervised manner for novel-view synthesis of dynamic unstructured scenes. Our end-to-end trainable algorithm learns highly complex, real-world static scenes within seconds and dynamic scenes with both rigid and non-rigid motion within minutes. By differentiating between static and motion-centric pixels, we create high-quality representations from a sparse set of images. We perform extensive qualitative and quantitative evaluation on existing benchmarks and set the state-of-the-art on performance measures on the challenging N…
Given a scene, what is going to move, and in what direction will it move? Such a question could be considered a non-semantic form of action prediction. In this work, we present a convolutional neural network (CNN) based approach for motion prediction. Given a static image, this CNN predicts the future motion of each and every pixel in the image in terms of optical flow. Our CNN model leverages the data in tens of thousands of realistic videos to train our model. Our method relies on absolutely no human labeling and is able to predict motion based on the context of the scene. Because our CNN mo…
The University of Hong Kong（香港大学）；University of Electronic Science and Technology of China（电子科技大学）；SenseTime（商汤科技）
Multi-agent motion prediction is challenging because it aims to foresee the future trajectories of multiple agents (\textit{e.g.} pedestrians) simultaneously in a complicated scene. Existing work addressed this challenge by either learning social spatial interactions represented by the positions of a group of pedestrians, while ignoring their temporal coherence (\textit{i.e.} dependencies between different long trajectories), or by understanding the complicated scene layout (\textit{e.g.} scene segmentation) to ensure safe navigation. However, unlike previous work that isolated the spatial int…
University of Toronto（多伦多大学）；Vector Institute（矢量研究所）；Snap Inc.（Snap公司）；CUHK（香港中文大学）；Stanford（斯坦福大学）；NTU（南洋理工大学）；HKU（香港大学）；University of Michigan（密歇根大学）；SFU（西蒙菲莎大学）；Google DeepMind（谷歌DeepMind）
Washington University（华盛顿大学）；Tsinghua University（清华大学）
This paper proposes an algorithm that turns a regular video capturing urban scenes into a high-quality endless animation, known as a Cinemagraph. The creation of a Cinemagraph usually requires a static camera in a carefully configured scene. The task becomes challenging for a regular video with a moving camera and objects. Our approach first warps an input video into the viewpoint of a reference camera. Based on the warped video, we propose effective temporal analysis algorithms to detect regions with static geometry and dynamic appearance, where geometric modeling is reliable and visually att…
The neuromorphic event cameras, which capture the optical changes of a scene, have drawn increasing attention due to their high speed and low power consumption. However, the event data are noisy, sparse, and nonuniform in the spatial-temporal domain with an extremely high temporal resolution, making it challenging to design backend algorithms for event-based vision. Existing methods encode events into point-cloud-based or voxel-based representations, but suffer from noise and/or information loss. Additionally, there is little research that systematically studies how to handle static and dynami…
Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；NVIDIA（英伟达）；MIT（麻省理工学院）
We address the unsupervised learning of several interconnected problems in low-level vision: single view depth prediction, camera motion estimation, optical flow, and segmentation of a video into the static scene and moving regions. Our key insight is that these four fundamental vision problems are coupled through geometric constraints. Consequently, learning to solve them together simplifies the problem because the solutions can reinforce each other. We go beyond previous work by exploiting geometry more explicitly and segmenting the scene into static and moving regions. To that end, we intro…
This work proposes a new end-to-end DCNN based approach for motion segmentation, especially for video sequences captured with such non-static cameras, called MOSNET. While other approaches focus on spatial or temporal context only, the proposed approach uses 3D convolutions as a key technology to factor in, spatio-temporal features in cohesive video frames. This is done by capturing temporal information in features with a low and also with a high level of abstraction. The lean network architecture with about 21k trainable parameters is mainly based on a pre-trained VGG-16 network. The MOSNET u…
Traditional grid/neighbor-based static pooling has become a constraint for point cloud geometry analysis. In this paper, we propose DAR-Net, a novel network architecture that focuses on dynamic feature aggregation. The central idea of DAR-Net is generating a self-adaptive pooling skeleton that considers both scene complexity and local geometry features. Providing variable semi-local receptive fields and weights, the skeleton serves as a bridge that connect local convolutional feature extractors and a global recurrent feature integrator. Experimental results on indoor scene datasets show advant…
In a given scene, humans can often easily predict a set of immediate future events that might happen. However, generalized pixel-level anticipation in computer vision systems is difficult because machine learning struggles with the ambiguity inherent in predicting the future. In this paper, we focus on predicting the dense trajectory of pixels in a scene, specifically what will move in the scene, where it will travel, and how it will deform over the course of one second. We propose a conditional variational autoencoder as a solution to this problem. In this framework, direct inference from the…
FreeDriveRF: Monocular RGB Dynamic NeRF without Poses for Autonomous Driving via Point-Level Dynamic-Static Decoupling
Department of Automation, Key Laboratory of System Control and Information Processing of Ministry of Education, Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education, Shanghai Engineering Research Center of Intelligent Control and Management（自动化系、教育部长系统控制与信息处理重点实验室、教育部长海洋智能设备与系统重点实验室、上海智能控制与管理工程研究中心）；Dimanshen Technology Co., Ltd.（迪马森科技有限公司）；Engineering Research Center of Intelligent Control for Underground Space, Ministry of Education, School of Information and Control Engineering, Advanced Robotics Research Center, China University of Mining and Technology（地下空间智能控制工程研究中心、教育部长、信息与控制工程学院、先进机器人研究中心、中国矿业大学）
