Datasets:

harpreetsahota
/

CarDD

@@ -58,167 +58,81 @@ dataset_summary: '
   '
 ---
-# Dataset Card for car_dd
-<!-- Provide a quick summary of the dataset. -->
-This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 2816 samples.
-## Installation
-If you haven't already, install FiftyOne:
-```bash
-pip install -U fiftyone
-```
-## Usage
-```python
-import fiftyone as fo
-from fiftyone.utils.huggingface import load_from_hub
-# Load the dataset
-# Note: other available arguments include 'max_samples', etc
-dataset = load_from_hub("harpreetsahota/CarDD")
-# Launch the App
-session = fo.launch_app(dataset)
-```
-## Dataset Details
-### Dataset Description
-<!-- Provide a longer summary of what this dataset is. -->
-- **Curated by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Language(s) (NLP):** en
-- **License:** [More Information Needed]
-### Dataset Sources [optional]
-<!-- Provide the basic links for the dataset. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-## Uses
-<!-- Address questions around how the dataset is intended to be used. -->
-### Direct Use
-<!-- This section describes suitable use cases for the dataset. -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
-[More Information Needed]
-## Dataset Structure
-<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
-[More Information Needed]
-## Dataset Creation
-### Curation Rationale
-<!-- Motivation for the creation of this dataset. -->
-[More Information Needed]
-### Source Data
-<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
-#### Data Collection and Processing
-<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
-[More Information Needed]
-#### Who are the source data producers?
-<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
-[More Information Needed]
-### Annotations [optional]
-<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
-#### Annotation process
-<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
-[More Information Needed]
-#### Who are the annotators?
-<!-- This section describes the people or systems who created the annotations. -->
-[More Information Needed]
-#### Personal and Sensitive Information
-<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
-[More Information Needed]
-## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
-## Citation [optional]
-<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
-**BibTeX:**
-[More Information Needed]
-**APA:**
-[More Information Needed]
-## Glossary [optional]
-<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
-[More Information Needed]
-## More Information [optional]
-[More Information Needed]
-## Dataset Card Authors [optional]
-[More Information Needed]
-## Dataset Card Contact
-[More Information Needed]

   '
 ---
+# 🚘 CarDD Dataset
+CarDD is a novel, public, large-scale dataset specifically designed for vision-based car damage detection and segmentation.
+The dataset contains **4,000 high-resolution car damage images** with over **9,000 well-annotated instances**, making it the largest public dataset of its kind.
+The high resolution of the images (average 684,231 pixels) is a key advantage over existing datasets that have a much lower average resolution (50,334 pixels). Higher resolution allows for more detailed annotations and the potential to detect finer damages.
+### CarDD Dataset Overview and Features
+CarDD features **six common external car damage categories**, chosen based on frequency of occurrence and clear definitions from insurance claim statistics.
+1. Dent
+2. Scratch
+3. Crack
+4. Glass shatter
+5. Tire flat
+6. Lamp broken
+### Annotation process
+The **annotation process** involved experts from the car insurance industry and trained annotators following specific guidelines based on insurance claim standards.
+These guidelines address challenges like
+• mixed damages (priority rules)
+• damages across components (boundary splitting)
+• adjacent same-class damages (boundary merging).
+For object detection and instance segmentation, the annotations include **masks and bounding boxes** associated with each of the six damage types.
+Each instance has a unique ID, category information, mask contours, and bounding box coordinates, following the COCO dataset format.
+For SOD, pixel-level binary ground truth maps are provided.
+### Dataset splits
+The dataset is split into **training (70.4%), validation (20.25%), and test (9.35%) sets**, maintaining a consistent ratio of instances for each category across the splits.
+Near-duplicate images were explicitly removed to prevent data leakage.
+### Uses
+The dataset provides **comprehensive annotations for multiple computer vision tasks**, including:
+    *   **Classification:** Identifying the type of damage.
+    *   **Object Detection:** Locating the damaged regions with bounding boxes.
+    *   **Instance Segmentation:** Precisely outlining the damaged areas with pixel-level masks.
+    *   **Salient Object Detection (SOD):** Identifying the damaged regions as salient objects through binary maps.
+CarDD presents several **challenges** for model development due to the nature of car damage:
+    *   **Fine-grained distinctions** between damage types like dents and scratches.
+    *   **Diversity in object scales and shapes** of the damages.
+    *   A **significant proportion of small objects**, particularly for dent, scratch, and crack categories.
+    *   The fact that damages like **dent, scratch, and crack can be intertwined and visually similar**.
+#### Availability
+The CarDD dataset is **publicly available** at https://cardd-ustc.github.io. However, access requires agreeing to the license terms of Flickr and Shutterstock, as the dataset does not own the copyright of the images. The dataset is intended for non-commercial research and educational purposes. Measures were taken to protect user privacy by mosaicking or deleting faces and license plates.
+# Citation
+```bibtex
+@ARTICLE{CarDD, author={Wang, Xinkuang and Li, Wenjing and Wu, Zhongcheng},
+journal={IEEE Transactions on Intelligent Transportation Systems},
+title={CarDD: A New Dataset for Vision-Based Car Damage Detection},
+year={2023},
+volume={24},
+number={7},
+pages={7202-7214},
+doi={10.1109/TITS.2023.3258480}}
+```