ToothFairy2
CBCT dataset of 480 volumes with fine-grained annotations spanning 42 maxillofacial classes (teeth, jaws, implants and more), in nnU-Net format with Hounsfield units. Released for the MICCAI 2024 ToothFairy2 challenge, it broadens the field of view and label set for automated dental segmentation.
ToothFairy3
CBCT dataset of 532 NIfTI volumes with comprehensive dental annotations covering 77 classes (teeth, jaws, canals and related structures) from multiple scanners. Released for the MICCAI 2025 ToothFairy3 challenge, it extends previous datasets for dental image segmentation.
ToothFairy
CBCT dataset for inferior alveolar nerve and dental segmentation, released for the MICCAI 2023 ToothFairy challenge. It provides 153 densely and 290 sparsely annotated volumes in NumPy format (Hounsfield units), improving on the earlier Maxillo dataset.
BarBeR
Barcode benchmarking dataset of 8,748 real images with 9,818 polygon-annotated barcodes across 18 1D and 2D symbologies (e.g. Code 128, EAN-13, QR, DataMatrix). Captured under varied conditions, it supports training and evaluation of barcode localization methods (ICPR 2024).
ToothFairy4
Multimodal dataset of 625 CBCT cases paired with clinical reports in Italian and English, from multiple institutions, for AI-generated dental and surgical-planning documentation. It is the benchmark for the ODIN 2026 challenge (MICCAI 2026), targeting assessment of dentition, bone quality and anatomical risks from 3D scans.
Pulpy3D
Dental CT dataset of 423 volumes with semantic and instance annotations across 19 classes, including teeth, inferior alveolar canals, dental pulp and anomalies. Provided as NIfTI files in Hounsfield units, it extends ToothFairy with pulp-focused labels for 3D pulp segmentation benchmarking (MICCAI 2024).
Bits2Bites
Dataset of 200 paired upper/lower intra-oral 3D scans (STL) of dental arches collected by Italian universities for occlusion analysis. The high-resolution meshes carry clinical annotations across sagittal, vertical and transverse relationships, enabling automated dental occlusion classification.
Maxillo
Maxillofacial CBCT dataset for inferior alveolar canal (IAC) segmentation, with 91 densely and 256 sparsely annotated 3D volumes produced by expert clinicians. It supports deep-learning research on mandibular canal detection and has since been superseded by the ToothFairy dataset.
Bite2Text
Orthodontic dataset of 2,000 patient cases pairing 3D intra-oral scans and 2D photographs with clinician-authored reports on malocclusion and treatment findings. Multi-center, with a hidden test set, it supports multimodal models that generate clinically meaningful orthodontic reports.
ResiDual
Industrial inspection dataset of 13,870 images grouped into 730 sequences of pharmaceutical vials under LED illumination, each showing clean vials or traces of rubber, glass, sand or plastic. Annotated across five classes, it targets computer-vision detection of particle contamination in pharmaceutical manufacturing.
TesticulUS
Collection of about 9,300 synthetic B-mode testicular ultrasound images (256x256 PNG) generated with denoising diffusion probabilistic models to address the scarcity of real clinical data. The synthetic images mimic real ultrasound appearance and noise, supporting medical image classification and pretraining.
No datasets match your search.