📜 BRIDGE-Open Dataset The BRIDGE-Open dataset is an open-access subset of the datasets used by the BRIDGE benchmark. This dataset includes the 55 tasks of the open-access datasets that can be released to the public. Each dataset should be used in accordance with the license of its original release (the source of each dataset is listed at our BRIDGE paper, Supplementary Section 5.)

Due to privacy and security considerations of clinical data, regulated-access datasets can not… See the full description on the dataset page: https://huggingface.co/datasets/YLab-Open/BRIDGE-Open.

Bridge Dataset 3 Manas

## Overview

Bridge Dataset 3 Manas is a dataset for object detection tasks - it contains Bridges A0JX annotations for 3,122 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [Public Domain license](https://creativecommons.org/licenses/Public Domain).

The Railroad Bridges dataset was compiled on October 14, 2022 from the Federal Railroad Administration (FRA) and is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation Atlas Database (NTAD). A railroad bridge is defined as “Railroad bridge means any structure with a deck, regardless of length, which supports one or more railroad tracks, or any other undergrade structure with an individual span length of 10 feet or more located at such a depth that it is affected by live loads.†based on the Code of Federal Regulations (49 CFR Part 237). The FRA does not have a mandate to inspect railroad bridges: these inspections are required by the owner of the track. The FRA will use this railroad bridge dataset to determine the number of bridges per railroad, state, etc. and will assist in determining priority field activities.

The National Bridge Inventory Elements dataset is as of June 20, 2025 from the Federal Highway Administration (FHWA) and is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation Statistics (BTS) National Transportation Atlas Database (NTAD). The data describes more than 615,000 of the Nation's bridges located on public roads, including Interstate Highways, U.S. highways, State and county roads, as well as publicly-accessible bridges on Federal and Tribal lands. The inventory data present a complete picture of the location, description, classification, and general condition data for each bridge. The element data present a breakdown of the condition of each structural and bridge management element for each bridge on the National Highway System (NHS). The Specification for the National Bridge Inventory Bridge Elements contains a detailed description of each data element including coding instructions and attribute definitions. The Coding Guide is available at: https://doi.org/10.21949/1519106. A data dictionary, or other source of attribute information, is accessible at https://doi.org/10.21949/1519105

jackyk02/bridge dataset hosted on Hugging Face and contributed by the HF Datasets community

Bridge

## Overview

Bridge is a dataset for object detection tasks - it contains Bridges annotations for 749 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [BY-NC-SA 4.0 license](https://creativecommons.org/licenses/BY-NC-SA 4.0).

This file contains location and identification information for bridges in Oregon. This includes bridges owned by the state, cities, counties, and other owners such as railroad bridges that cross state highways. It does not include bridges that are owned by Federal agencies. A bridge is a structure including supports erected over a depression or an obstruction, such as water, highway, or railway, and having a track or passageway for carrying traffic or other moving loads and having an opening measured along the center of the roadway of more than 20 feet.Additional metadata resouce: https://geoportalprod-ordot.msappproxy.net/geoportal/catalog/main/home.page

The Bridge dataset was developed using data provided by the Bridge Inspection Database. The Bridge Inspection Database contains a record for each Bridge Structure on public roadways in Texas. This includes Bridges maintained by TxDOT, Toll Authorities, Counties, Municipalities, and other jurisdictions. Bridge Inspection data is the primary source used to update the National Bridge Inspection File (NBI) in Washington. For more information on Bridge attribute data, please consult the Bridge Inspection Coding Guide and the Bridge Data Dictionary found at the links below:TxDOT Bridge Coding GuideTxDOT Bridge Data DictionaryUpdate Frequency: NightlySecurity Level: Public

Bridges are inspected every two years at minimum; bridges in poor condition are inspected more frequently. These inspections rate the conditions of various bridge elements. The inspection records are stored in a bridge management system. The Federal Highway Administration annually collects data on bridges more than 20 feet long from each state’s department of transportation, including the Pennsylvania Department of Transportation and New Jersey Department of Transportation, and stores it in their National Bridge Inventory.

This analysis considers a bridge's deck, super, and substructure ratings, or the culvert rating for a culvert. A score between 0 and 9 is given for each component. Bridges with scores between 7 to 9 for all three components are rated "good," those with any component scores of 4 or lower are rated "poor." A bridge that doesn't fall into a good or poor rating is considered “fair”.

Most bridges are on state roads and are maintained by the state’s department of transportation. Some bridges are on local roads and are maintained by local governments. The "Other" category covers bridges maintained by turnpike and toll authorities.

P.bridge

## Overview

P.bridge is a dataset for object detection tasks - it contains Car annotations for 948 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

The NBI is a collection of information (database) describing the more than 600,000 of the Nation's bridges located on public roads, including Interstate Highways, U.S. highways, State and county roads, as well as publicly-accessible bridges on Federal lands. It presents a State by State summary analysis of the number, location, and general condition of highway bridges within each State.

"Image.zip" contains 955 corrrosion images, 1480 crack images, 1269 free lime images, 873 water leakage images, and 1244 spalling images. These images are labeled with numbers from 0 to 6 including the background. The "Label.zip" file contains the labeled images, and the "Image.json" file contains the label information.

Homeland Infrastructure Foundation-Level Data (HIFLD) geospatial data sets containing information on Bridge Areas (USACE IENC).

Bridge Damage Detection

## Overview

Bridge Damage Detection is a dataset for object detection tasks - it contains Damage annotations for 3,132 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This inventory of Detroit pedestrian bridges and their condition assessments is derived from the Michigan Department of Transportation (MDOT) and includes pedestrian bridges that span State of Michigan roadways only.Inspections of pedestrian bridges over State of Michigan routes are performed on a risk-based interval by MDOT, not to exceed 24 months. While not governed by the National Bridge Inspection Standards, these inspections are performed to maintain safety of the traveled way and to provide information needed for asset management. Routine Inspections of the entire structure may be supplemented by Special Inspections of Bridge Components that require more frequent monitoring. Condition ratings for pedestrian bridges are evaluated following the National Bridge Inventory Condition Ratings in a similar manner to vehicle bridges. These ratings are based on a 0-9 scale, where 0 is "Failed condition" and 9, "Excellent condition". Condition ratings are assigned for the deck, superstructure and substructure of each bridge or for each culvert. A culvert is a structure that allows passage under a roadway and has a short span (less than 20 feet).

Vietnam Westside Bridge

## Overview

Vietnam Westside Bridge is a dataset for object detection tasks - it contains Bridge Watter annotations for 630 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Segmenting Bridge Elements

## Overview

Segmenting Bridge Elements is a dataset for object detection tasks - it contains Bridge Elements annotations for 347 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Bridges-Rail in the United States According to The National Bridge Inspection Standards published in the Code of Federal Regulations (23 CFR 650.3), a bridge is: A structure including supports erected over a depression or an obstruction, such as water, highway, or railway, and having a track or passageway for carrying traffic or other moving loads. Each bridge was captured as a point which was placed in the center of the "main span" (highest and longest span). For bridges that cross navigable waterways, this was typically the part of the bridge over the navigation channel. If no "main span" was discernable using the imagery sources available, or if multiple non contiguous main spans were discernable, the point was placed in the center of the overall structure. Bridges that are sourced from the National Bridge Inventory (NBI) that cross state boundaries are an exception. Bridges that cross state boundaries are represented in the NBI by two records. The points for the two records have been located so as to be within the state indicated by the NBI's [STATE_CODE] attribute. In some cases, following these rules did not place the point at the location at which the bridge crosses what the user may judge as the most important feature intersected. For example, a given bridge may be many miles long, crossing nothing more than low lying ground for most of its length but crossing a major interstate at its far end. Due to the fact that bridges are often high narrow structures crossing depressions that may or may not be too narrow to be represented in the DEM used to orthorectify a given source of imagery, alignment with ortho imagery is highly variable. In particular, apparent bridge location in ortho imagery is highly dependent on collection angle. During verification, TechniGraphics used imagery from the following sources: NGA HSIP 133 City, State or Local; NAIP; DOQQ imagery. In cases where "bridge sway" or "tall structure lean" was evident, TGS attempted to compensate for these factors when capturing the bridge location. For instances in which the bridge was not visible in imagery, it was captured using topographic maps at the intersection of the water and rail line. TGS processed 784 entities previously with the HSIP Bridges-Roads (STRAHNET Option - HSIP 133 Cities and Gulf Coast). These entities were added into this dataset after processing. No entities were included in this dataset for American Samoa, Guam, Hawaii, the Commonwealth of the Northern Mariana Islands, or the Virgin Islands because there are no main line railways in these areas. At the request of NGA, text fields in this dataset have been set to all upper case to facilitate consistent database engine search results. At the request of NGA, leading and trailing spaces were trimmed from all text fields. At the request of NGA, all diacritics (e.g., the German umlaut or the Spanish tilde) have been replaced with their closest equivalent English character to facilitate use with database systems that may not support diacritics. The currentness of this dataset is given by the publication date which is 09/02/2009. A more precise measure of currentness cannot be provided since this is dependent on the NBI and the source of imagery used during processing.

Retrieval-Augmented Generation (RAG) Dataset 1200

Retrieval-Augmented Generation (RAG) Dataset 1200 is an English dataset designed for RAG-optimized models, built by Neural Bridge AI, and released under Apache licence 2.0.

  Dataset Description





  Dataset Summary

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by allowing them to consult an external authoritative knowledge base before generating responses. This approach significantly… See the full description on the dataset page: https://huggingface.co/datasets/neural-bridge/rag-dataset-1200.

All bridges in Ontario are inspected every 2 years using the Ontario Structure Inspection Manual (OSIM) format. The weighted condition of all elements is summarized in the Bridge Condition Index (BCI). More details about the BCI can be found on the Bridge Safety page on the Ministry of Transportation website.

Bridges that were recently constructed or recently transferred to the Ministry of Transportation may not have been inspected in OSIM format yet.

*[OSIM]: Ontario Structure Inspection Manual *[BCI]: Bridge Condition Index