An official website of the United States government
Existing approaches for autonomous control of pan-tilt-zoom (PTZ) cameras use multiple stages where object detection and localization are performed separately from the control of the PTZ mechanisms. These approaches require manual labels and suffer from performance bottlenecks due to error propagation across the multi-stage flow of information. The large size of object detection neural networks also makes prior solutions infeasible for real-time deployment in resource-constrained devices. We present an end-to-end deep reinforcement learning (RL) solution called Eagle1 to train a neural network policy that directly takes images as input to control the PTZ camera. Training reinforcement learning is cumbersome in the real world due to labeling effort, runtime environment stochasticity, and fragile experimental setups. We introduce a photo-realistic simulation framework for training and evaluation of PTZ camera control policies. Eagle achieves superior camera control performance by maintaining the object of interest close to the center of captured images at high resolution and has up to 17% more tracking duration than the state-of-the-art. Eagle policies are lightweight (90x fewer parameters than Yolo5s) and can run on embedded camera platforms such as Raspberry PI (33 FPS) and Jetson Nano (38 FPS), facilitating real-time PTZ tracking for resource-constrained environments. With domain randomization, Eagle policies trained in our simulator can be transferred directly to real-world scenarios2.
more »
« less
What Does Learning About Time Tell About Outdoor Scenes?
In this paper, we explore the potential of utilizing time-stamps as labels for Deep Learning from webcams, surveillance cameras, and other fixed viewpoint image situations. Specifically, we explore if learning to classify images by the time they were taken uncovers interesting patterns and behaviors in the scenes captured by these cameras. We describe approaches to building datasets with large quantities of images and their accompanying labels, making them suitable for large-scale deep learning approaches. We share our results from the initial deep learning experiments.
more »
« less
- Award ID(s):
- 2125677
- PAR ID:
- 10448619
- Date Published:
- Journal Name:
- Proceedings of the 2022 IEEE Applied Imagery Pattern Recognition Workshop (AIPR)
- Page Range / eLocation ID:
- 1 to 6
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Accurate cloud type identification and coverage analysis are crucial in understanding the Earth’s radiative budget. Traditional computer vision methods rely on low-level visual features of clouds for estimating cloud coverage or sky conditions. Several handcrafted approaches have been proposed; however, scope for improvement still exists. Newer deep neural networks (DNNs) have demonstrated superior performance for cloud segmentation and categorization. These methods, however, need expert engineering intervention in the preprocessing steps—in the traditional methods—or human assistance in assigning cloud or clear sky labels to a pixel for training DNNs. Such human mediation imposes considerable time and labor costs. We present the application of a new self-supervised learning approach to autonomously extract relevant features from sky images captured by ground-based cameras, for the classification and segmentation of clouds. We evaluate a joint embedding architecture that uses self-knowledge distillation plus regularization. We use two datasets to demonstrate the network’s ability to classify and segment sky images—one with ∼ 85,000 images collected from our ground-based camera and another with 400 labeled images from the WSISEG database. We find that this approach can discriminate full-sky images based on cloud coverage, diurnal variation, and cloud base height. Furthermore, it semantically segments the cloud areas without labels. The approach shows competitive performance in all tested tasks,suggesting a new alternative for cloud characterization.more » « less
-
It is a common practice to think of a video as a sequence of images (frames), and re-use deep neural network models that are trained only on images for similar analytics tasks on videos. In this paper, we show that this “leap of faith” that deep learning models that work well on images will also work well on videos is actually flawed.We show that even when a video camera is viewing a scene that is not changing in any humanperceptible way, and we control for external factors like video compression and environment (lighting), the accuracy of video analytics application fluctuates noticeably. These fluctuations occur because successive frames produced by the video camera may look similar visually, but are perceived quite differently by the video analytics applications.We observed that the root cause for these fluctuations is the dynamic camera parameter changes that a video camera automatically makes in order to capture and produce a visually pleasing video. The camera inadvertently acts as an “unintentional adversary” because these slight changes in the image pixel values in consecutive frames, as we show, have a noticeably adverse impact on the accuracy of insights from video analytics tasks that re-use image-trained deep learning models. To address this inadvertent adversarial effect from the camera, we explore the use of transfer learning techniques to improve learning in video analytics tasks through the transfer of knowledge from learning on image analytics tasks. Our experiments with a number of different cameras, and a variety of different video analytics tasks, show that the inadvertent adversarial effect from the camera can be noticeably offset by quickly re-training the deep learning models using transfer learning. In particular, we show that our newly trained Yolov5 model reduces fluctuation in object detection across frames, which leads to better tracking of objects (∼40% fewer mistakes in tracking). Our paper also provides new directions and techniques to mitigate the camera’s adversarial effect on deep learning models used for video analytics applications.more » « less
-
Deep learning algorithms are exceptionally valuable tools for collecting and analyzing the catastrophic readiness and countless actionable flood data. Convolutional neural networks (CNNs) are one form of deep learning algorithms widely used in computer vision which can be used to study flood images and assign learnable weights to various objects in the image. Here, we leveraged and discussed how connected vision systems can be used to embed cameras, image processing, CNNs, and data connectivity capabilities for flood label detection. We built a training database service of >9000 images (image annotation service) including the image geolocation information by streaming relevant images from social media platforms, Department of Transportation (DOT) 511 traffic cameras, the US Geological Survey (USGS) live river cameras, and images downloaded from search engines. We then developed a new python package called “FloodImageClassifier” to classify and detect objects within the collected flood images. “FloodImageClassifier” includes various CNNs architectures such as YOLOv3 (You look only once version 3), Fast R–CNN (Region-based CNN), Mask R–CNN, SSD MobileNet (Single Shot MultiBox Detector MobileNet), and EfficientDet (Efficient Object Detection) to perform both object detection and segmentation simultaneously. Canny Edge Detection and aspect ratio concepts are also included in the package for flood water level estimation and classification. The pipeline is smartly designed to train a large number of images and calculate flood water levels and inundation areas which can be used to identify flood depth, severity, and risk. “FloodImageClassifier” can be embedded with the USGS live river cameras and 511 traffic cameras to monitor river and road flooding conditions and provide early intelligence to emergency response authorities in real-time.more » « less
-
null (Ed.)Lensless imaging is a new, emerging modality where image sensors utilize optical elements in front of the sensor to perform multiplexed imaging. There have been several recent papers to reconstruct images from lensless imagers, including methods that utilize deep learning for state-of-the-art performance. However, many of these methods require explicit knowledge of the optical element, such as the point spread function, or learn the reconstruction mapping for a single fixed PSF. In this paper, we explore a neural network architecture that performs joint image reconstruction and PSF estimation to robustly recover images captured with multiple PSFs from different cameras. Using adversarial learning, this approach achieves improved reconstruction results that do not require explicit knowledge of the PSF at test-time and shows an added improvement in the reconstruction model’s ability to generalize to variations in the camera’s PSF. This allows lensless cameras to be utilized in a wider range of applications that require multiple cameras without the need to explicitly train a separate model for each new camera.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/AIPR57179.2022.10092235
