DukeUltrasound is an ultrasound dataset collected at Duke University with a Verasonics c52v probe. It contains delay-and-sum (DAS) beamformed data as well as data post-processed with Siemens Dynamic TCE for speckle reduction, contrast enhancement and improvement in conspicuity of anatomical structures. These data were collected with support from the National Institute of Biomedical Imaging and Bioengineering under Grant R01-EB026574 and National Institutes of Health under Grant 5T32GM007171-44. A usage example is available here.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('duke_ultrasound', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

Introduction Generation customers connected to UK Power Networks can be subjected to curtailment through our Distributed Energy Resource Management System (DERMS) if they accepted a curtailable-connection. During periods of network congestion, these DERS will have their access reduced to mitigate network constraint breaches. Their reduction is organised according to their connection application date in a last-in first-out (LIFO) arrangement. The Constraints Real Time Meter Readings dataset on the Open Data Portal (ODP) gives a near real time status of the constraints on our network that are used by DERMS to reduce access. This API accessible dataset can be used to see just how congested the network is, and for the specific DER operators themselves, they have access and visibility to the constraints of their specific site. The dataset contains a timestamp, the constraint identifier, the most recent current reading in amps, the trim and release limits (curtailment starts at the trim and ends at the release), whether the site is in breach, a description of the constraint, and (only if you have access) the name of the DER. The dataset updates as close to real time as is possible. Our scheduling is as follows:

At 15s past the minute mark, we scrape the network data and push it to the ODP server On the minute mark, the ODP runs an update to refresh the dataset The dataset refresh is completed between 5-15s past the minute mark Only after this refresh has completed can you get the latest values from the ODP

You can run this notebook to see the dataset in action: https://colab.research.google.com/drive/1Czx98U6zttlA3PC2OfI_0UzAbE48BvEq?usp=sharing

Methodological Approach

A Remote Terminal Unit (RTU) is installed at each curtailable-connection site providing live telemetry data into the DERMS. It measures communications status, generator output, and mode of operation. RTUs are also installed at constraint locations (physical parts of the network, e.g., transformers, cables which may become overloaded under certain conditions). These are identified through planning power load studies. These RTUs monitor current at the constraint and communications status. The DERMS design integrates network topology information. This maps constraints to associated curtailable connections under different network running conditions, including the sensitivity of the constraints to each curtailable connection. In general, a 1MW reduction in generation of a customer will cause <1MW reduction at the constraint. Each constraint is registered to a GSP. DERMS monitors constraints against the associated breach limit. When a constraint limit is breached, DERMS calculates the amount of access reduction required from curtailable connections linked to the constraint to alleviate the breach. This calculation factors in the real-time level of generation of each customer and the sensitivity of the constraint to each generator. Access reduction is issued to each curtailable-connection via the RTU until the constraint limit breach is mitigated. Multiple constraints can apply to a curtailable-connection and constraint breaches can occur simultaneously. Where multiple constraint breaches act upon a single curtailable-connection, we apportion the access reduction of that connection to the constraint breaches depending on the relative magnitude of the breaches. Where customer curtailment occurs without any associated constraint breach, we categorize the curtailment as non-constraint driven. Future developments will include the reason for non-constraint driven curtailment.

Quality Control Statement Quality Control Measures include:

Manual review and correction of data inconsistencies. Use of additional verification steps to ensure accuracy in the methodology.

Assurance Statement The DSO Data Science Team checked to ensure data accuracy and consistency.

Other Download dataset information: Metadata (JSON) Definitions of key terms related to this dataset can be found in the Open Data Portal Glossary: https://ukpowernetworks.opendatasoft.com/pages/glossary/

This work has been published in the Nature Scientific Data. Suggested citation: Rajib et al. The changing face of floodplains in the Mississippi River Basin detected by a 60-year land use change dataset. Nature Scientific Data 8, 271 (2021). https://doi.org/10.1038/s41597-021-01048-w

Here, we present the first-available dataset that quantifies land use change along the floodplains of the Mississippi River Basin (MRB) covering 60 years (1941-2000) at 250-m resolution. The MRB is the fourth largest river basin in the world (3.3 million sq km) comprising 41% of the United States and draining into the Gulf of Mexico, an area with an annually expanding and contracting hypoxic zone resulting from basin-wide over-enrichment of nutrients. The basin represents one of the most engineered systems in the world, and includes complex web of dams, levees, floodplains, and dikes. This new dataset reveals the heterogenous spatial extent of land use transformations in MRB floodplains. The domination transition of floodplains has been from natural ecosystems (e.g. wetlands or forests) to agricultural use. A steady increase in developed land use within the MRB floodplains was also evident.

To maximize the reuse of this dataset, our contributions also include four unique products: (i) a Google Earth Engine interactive map visualization interface: https://gishub.org/mrb-floodplain (ii) a Google-based Python code that runs in any internet browser: https://colab.research.google.com/drive/1vmIaUCkL66CoTv4rNRIWpJXYXp4TlAKd?usp=sharing (iii) an online tutorial with visualizations facilitating classroom application of the code: https://serc.carleton.edu/hydromodules/steps/241489.html (iv) an instructional video showing how to run the code and partially reproduce the floodplain land use change dataset: https://youtu.be/wH0gif_y15A

Guanaco: Lazy Llama 2 Formatting

This is the excellent timdettmers/openassistant-guanaco dataset, processed to match Llama 2's prompt format as described in this article. Useful if you don't want to reformat it by yourself (e.g., using a script). It was designed for this article about fine-tuning a Llama 2 model in a Google Colab.

ALL CONTENT IS DIRECTLY FROM THE GITHUB README

RF-DETR: SOTA Real-Time Object Detection Model

• hf space link
• colab demo link
• roboflow blog link
• discord link

RF-DETR is a real-time, transformer-based object detection model architecture developed by Roboflow and released under the Apache 2.0 license.

RF-DETR is the first real-time model to exceed 60 AP on the Microsoft COCO benchmark alongside competitive performance at base sizes. It also achieves state-of-the-art performance on RF100-VL, an object detection benchmark that measures model domain adaptability to real world problems. RF-DETR is comparable speed to current real-time objection models.

RF-DETR is small enough to run on the edge, making it an ideal model for deployments that need both strong accuracy and real-time performance.

Results

We validated the performance of RF-DETR on both Microsoft COCO and the RF100-VL benchmarks.

https://media.roboflow.com/rf-detr/charts.png" alt="rf-detr-coco-rf100-vl-8">

| Model | mAP