datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
IDBench-Omni
IDBench-Omni
IDBench-Omni is a benchmark for controllable human-centric audio-video generation. It contains three tasks:
Task
Subsets
Samples
Inputs
Target
R2AV
single_person, multi_person
100
text prompt, reference identity image(s), reference voice audio(s)
generate synchronized video and audio
RA2V
default
50
text prompt, reference identity image, driving audio
animate the identity with the driving audio
RV2AV
swap_face, swap_human
50
text prompt, reference… See the full description on the dataset page: https://huggingface.co/datasets/XuGuo699/IDBench-Omni.ALL-IDB-Patches
ALL-IDB Patches
MATLAB source code for creating image patches and labels used in the paper “ALL-IDB Patches: Whole slide imaging for Acute Lymphoblastic Leukemia detection using Deep Learning”, presented at ICASSP Workshops 2023.
The repository converts annotated ALL-IDB1 whole-slide microscope images into fixed-size overlapping patches, preserving the position of white blood cell centroids and generating patch-level labels for probable lymphoblasts… See the full description on the dataset page: https://huggingface.co/datasets/AngeloUNIMI/ALL-IDB-Patches.ID-Bench
ID-Bench
ID-Bench is a real-world benchmark for multi-reference identity-preserving image generation. It is built from real-world e-commerce advertising images and organized by product identity, with the goal of evaluating whether a model can generate a novel target image that both preserves product identity and follows target-specific variation cues.
We release, in this repository, the curated benchmark dataset that is consistent with the one used for evaluation in our paper. The… See the full description on the dataset page: https://huggingface.co/datasets/zyyyz/ID-Bench.idbi_dataset