An official website of the United States government
In recent years, novel view synthesis from a single image has seen significant progress thanks to the rapid advancements in 3D scene representation and image inpainting techniques. While the current approaches are able to synthesize geometrically consistent novel views, they often do not handle the view-dependent effects properly. Specifically, the highlights in their synthesized images usually appear to be glued to the surfaces, making the novel views unrealistic. To address this major problem, we make a key observation that the process of synthesizing novel views requires changing the shading of the pixels based on the novel camera, and moving them to appropriate locations. Therefore, we propose to split the view synthesis process into two independent tasks of pixel reshading and relocation. During the reshading process, we take the single image as the input and adjust its shading based on the novel camera. This reshaded image is then used as the input to an existing view synthesis method to relocate the pixels and produce the final novel view image. We propose to use a neural network to perform reshading and generate a large set of synthetic input-reshaded pairs to train our network. We demonstrate that our approach produces plausible novel view images with realistic moving highlights on a variety of real world scenes.
more »
« less
SimAug: Learning Robust Representations from Simulation for Trajectory Prediction
This paper focuses on the problem of predicting future trajectories of people in unseen scenarios and camera views. We propose a method to efficiently utilize multi-view 3D simulation data for training. Our approach finds the hardest camera view to mix up with adversarial data from the original camera view in training, thus enabling the model to learn robust representations that can generalize to unseen camera views. We refer to our method as SimAug. We show that SimAug achieves best results on three out-of-domain real-world benchmarks, as well as getting state-of-the-art in the Stanford Drone and the VIRAT/ActEV dataset with in-domain training data. We will release our models and code.
more »
« less
- Award ID(s):
- 1650994
- PAR ID:
- 10289118
- Editor(s):
- Vedaldi, Andrea; Bischof, Horst; Brox, Thomas; Frahm, Jan-Michael
- Date Published:
- Journal Name:
- Computer Vision – ECCV 2020
- Volume:
- 12358
- Issue:
- 978-3-030-58601-0
- Page Range / eLocation ID:
- 275-292
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Human activity recognition provides insights into physical and mental well-being by monitoring patterns of movement and behavior, facilitating personalized interventions and proactive health management. Radio Frequency (RF)-based human activity recognition (HAR) is gaining attention due to its less privacy exposure and non-contact characteristics. However, it suffers from data scarcity problems and is sensitive to environment changes. Collecting and labeling such data is laborintensive and time consuming. The limited training data makes generalizability challenging when the sensor is deployed in a very different relative view in the real world. Synthetic data generation from abundant videos presents a potential to address data scarcity issues, yet the domain gaps between synthetic and real data constrain its benefit. In this paper, we firstly share our investigations and insights on the intrinsic limitations of existing video-based data synthesis methods. Then we present M4X, a method using metric learning to extract effective viewindependent features from the more abundant synthetic data despite their domain gaps, thus enhancing cross-view generalizability. We explore two main design issues in different mining strategies for contrastive pairs/triplets construction, and different forms of loss functions. We find that the best choices are offline triplet mining with real data as anchors, balanced triplets, and a triplet loss function without hard negative mining for higher discriminative power. Comprehensive experiments show that M4X consistently outperform baseline methods in cross-view generalizability. In the most challenging case of the least amount of real training data, M4X outperforms three baselines by 7.9- 16.5% on all views, and 18.9-25.6% on a view with only synthetic but no real data during training. This proves its effectiveness in extracting view-independent features from synthetic data despite their domain gaps. We also observe that given limited sensor deployments, a participant-facing viewpoint and another at a large angle (e.g. 60◦) tend to produce much better performance.more » « less
-
An organ segmentation method that can generalize to unseen contrasts and scanner settings can significantly reduce the need for retraining of deep learning models. Domain Generalization (DG) aims to achieve this goal. However, most DG methods for segmentation require training data from multiple domains during training. We propose a novel adversarial domain generalization method for organ segmentation trained on data from a single domain. We synthesize the new domains via learning an adversarial domain synthesizer (ADS) and presume that the synthetic domains cover a large enough area of plausible distributions so that unseen domains can be interpolated from synthetic domains. We propose a mutual information regularizer to enforce the semantic consistency between images from the synthetic domains, which can be estimated by patch-level contrastive learning. We evaluate our method for various organ segmentation for unseen modalities, scanning protocols, and scanner sites.more » « less
-
In computer vision, tracking humans across camera views remains challenging, especially for complex scenarios with frequent occlusions, significant lighting changes and other difficulties. Under such conditions, most existing appearance and geometric cues are not reliable enough to distinguish humans across camera views. To address these challenges, this paper presents a stochastic attribute grammar model for leveraging complementary and discriminative human attributes for enhancing cross-view tracking. The key idea of our method is to introduce a hierarchical representation, parse graph, to describe a subject and its movement trajectory in both space and time domains. This results in a hierarchical compositional representation, comprising trajectory entities of varying level, including human boxes, 3D human boxes, tracklets and trajectories. We use a set of grammar rules to decompose a graph node (e.g. tracklet) into a set of children nodes (e.g. 3D human boxes), and augment each node with a set of attributes, including geometry (e.g., moving speed, direction), accessories (e.g., bags), and/or activities (e.g., walking, running). These attributes serve as valuable cues, in addition to appearance features (e.g., colors), in determining the associations of human detection boxes across cameras. In particular, the attributes of a parent node are inherited by its children nodes, resulting in consistency constraints over the feasible parse graph. Thus, we cast cross-view human tracking as finding the most discriminative parse graph for each subject in videos. We develop a learning method to train this attribute grammar model from weakly supervised training data. To infer the optimal parse graph and its attributes, we develop an alternative parsing method that employs both top-down and bottom-up computations to search the optimal solution. We also explicitly reason the occlusion status of each entity in order to deal with significant changes of camera viewpoints. We evaluate the proposed method over public video benchmarks and demonstrate with extensive experiments that our method clearly outperforms state-of-theart tracking methods.more » « less
-
Multi-camera systems are indispensable in movies, TV shows, and other media. Selecting the appropriate camera at every timestamp has a decisive impact on production quality and audience preferences. Learning-based view recommendation frameworks can assist professionals in decision-making. However, they often struggle outside of their training domains. The scarcity of labeled multi-camera view recommendation datasets exacerbates the issue. Based on the insight that many videos are edited from the original multi-camera videos, we propose transforming regular videos into pseudo-labeled multi-camera view recommendation datasets. Promisingly, by training the model on pseudo-labeled datasets stemming from videos in the target domain, we achieve a 68% relative improvement in the model’s accuracy in the target domain and bridge the accuracy gap between in-domain and never-before-seen domains.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/0.1007/978-3-030-58601-0_17
