Papers with Code 아카이브

Image Generation

태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.

하위 태스크
23
평가표
93
보존된 평가 기록
870

Image Generation (synthesis) is the task of generating new images from an existing dataset.

  • Unconditional generation refers to generating samples unconditionally from the dataset, i.e. $p(y)$
  • Conditional image generation (subtask) refers to generating samples conditionally from the dataset, based on a label, i.e. $p(y|x)$.

In this section, you can find state-of-the-art leaderboards for unconditional generation. For conditional generation, and other types of image generations, refer to the subtasks.

<span style="color:grey; opacity: 0.6">( Image credit: StyleGAN )</span>

데이터셋

출처에 같은 태스크 이름으로 연결된 데이터셋입니다. 동명 태스크가 포함될 수 있으므로, 비교할 때는 평가 결과의 조건도 확인하세요.

BDD100K-Subsets

Subsets of BDD100K Dataset that are used in Object Detection Under Rainy Conditions for Autonomous Vehicles: A Review of State-of-the-Art and Emerging Techniques

Cam2BEV

The dataset contains two subsets of synthetic, semantically segmented road-scene images, which have been created for developing and applying the methodology described in the paper "A Sim2Real Deep Learning Approach for the Transformation of Images from Multiple Vehicle-Mounted Cameras to a Semantically Segmented Image in Bird’s Eye View" (IEEE Xplore, arXiv, YouTube) The dataset can be used through the official code implementation of the Cam2BEV methodology described on Github. | Dataset | # Training Samples | # Validation Samples | # Vehicle Cameras | # Semantic Classes | Contained Images (examples) | | --- | --- | --- | --- | --- | --- | | Dataset 1: 360° Surround | 33199 | 3731 | 4 (front, rear, left, right) | 30 (CityScapes) | front camera, rear camera, left camera, right camera, bird's eye view, [bird's eye view incl. occlusion](https://gitlab.ika.rwth-aachen.de/cam2bev/cam2bev-data/-/raw/master/1_FRLR/examples bev+occlusion.png), homography view | | Dataset 2: Front Camera only | 32246 | 3172 | 1 (front) | 30 (CityScapes) | front camera, bird's eye view, bird's eye view incl. occlusion, homography view |

E.T. the Exceptional Trajectories

Click to add a brief description of the dataset (Markdown and LaTeX enabled). Provide: * a high-level explanation of the dataset characteristics * explain motivations and summary of its content * potential use cases of the dataset

ENTIGEN

ENTIGEN is a benchmark dataset to evaluate the change in image generations conditional on ethical interventions across three social axes -- gender, skin color, and culture. It contains 246 prompts based on an attribute set containing diverse professions, objects, and cultural scenarios.

Evalmuse-40k

Recently, Text-to-Image (T2I) generation models have achieved significant advancements. Correspondingly, many automated metrics have emerged to evaluate the image-text alignment capabilities of generative models. However, the performance comparison among these automated metrics is limited by existing small datasets. Additionally, these datasets lack the capacity to assess the performance of automated metrics at a fine-grained level. In this study, we contribute an EvalMuse-40K benchmark, gathering 40K image-text pairs with fine-grained human annotations for image-text alignment-related tasks. In the construction process, we employ various strategies such as balanced prompt sampling and data re-annotation to ensure the diversity and reliability of our benchmark.

Mila Simulated Floods

Mila Simulated Floods Dataset is a 1.5 square km virtual world using the Unity3D game engine including urban, suburban and rural areas. The urban environment contains skyscrapers, large buildings, and roads, as well as objects such as traffic items and vehicles. The rural environment consists of a landscape of grassy hills, forests, and mountains, with sparse houses and other buildings such as a church, and no roads. The rural and urban areas make up for 1 square km of our virtual world. The suburban environment is a residential area of 0.5 square km with many individual houses with front yards. To gather the simulated dataset, we captured before and after flood pairs from 2000 viewpoints with the following modalities: - non-flooded RGB image, depth map, segmentation map - flooded RGB image, binary mask of the flooded area, segmentation map The camera was placed about 1.5 above ground, and has a field of view of 120 degree, and the resolution of the images is 1200 x 900. At each viewpoint, we took 10 pictures, by varying slightly the position of the camera in order to augment the dataset.

OADAT

An experimental and synthetic (simulated) OA raw signals and reconstructed image domain datasets rendered with different experimental parameters and tomographic acquisition geometries. For detailed information, see github.com/berkanlafci/oadat.

PAL4Inpaint

PAL4Inpaint is a dataset consisting of 4,795 inpainting results with per-pixel perceptual artifacts annotations designed for image inpainting tasks.

Paper2Fig100k

Paper2Fig100k is a dataset with over 100k images of figures and texts from research papers. The figures show architecture diagrams and methodologies of articles available at arXiv.org from fields like artificial intelligence and computer vision. Figures usually include text and discrete objects, e.g., boxes in a diagram, with lines and arrows that connect them.

RichHF-18K

We collect a dataset of Rich Human Feedback on 18K images (RichHF-18K), which contains (i) point annotations on the image that highlight regions of implausibility/artifacts, and text-image misalignment; (ii) labeled words on the prompts specifying the missing or misrepresented concepts in the generated image; and (iii) four types of fine-grained scores for image plausibility, text-image alignment, aesthetics, and overall rating.

2007 BP Anisotropic Velocity Benchmark

The 2007 BP TTI Velocity-Analysis Benchmark dataset was created by Hemang Shah and provided courtesy of BP Exploration Operation Company Limited ("BP"). It was originally used at the velocity-analysis workshop at the 2010 EAGE meeting in Barcelona, Spain.

Apolloscape Inpainting

The Inpainting dataset consists of synchronized Labeled image and LiDAR scanned point clouds. It's captured by HESAI Pandora All-in-One Sensing Kit. It is collected under various lighting conditions and traffic densities in Beijing, China.

Boombox

Boombox is a multi-modal dataset for visual reconstruction from acoustic vibrations. Involves dropping objects into a box and capturing resulting images and vibrations. Used for training ML systems that predict images from vibration. Potential application domain: Computer Vision, Multimodal Perception, Vision and Sound, Sight from Sound, Robotics, Deep Learning, and Machine Learning.

CARLA2Real

This dataset contains synthetic images extracted from the CARLA simulator along with rich information extracted from the deferred rendering pipeline of Unreal Engine 4. The main purpose of this dataset is the training of the state-of-the-art image-to-image translation model proposed by Intel Labs "Enhancing Photorealism Enhancement" (EPE). Translation results derived from the model targeting the characteristics of Cityscapes, KITTI, and Mapillary Vistas are also provided. Computer vision-based models trained on these data are expected to perform better when deployed in the real world.

Colors

A large dataset of color names and their respective RGB values stores in CSV.

CPM-Synt-1

CPM-Synt-1 is a dataset consisting of 5555 images with synthesis - makeup images with pattern segmentation mask

CPM-Synt-2

CPM-Synt-2 is a dataset consisting of 1625 images with synthesis - triplets: makeup, non-makeup, ground-truth.

FFHQH

A new dataset for portrait harmonization based on the FFHQ. It contains real images, foreground masks, and synthesized composites. Generating composite images has two steps: alpha mask generation and color augmentation. StyleMatte - portrait matting network was used for mask generation. Color augmentations: brightness, contrast, hue, sharpness adjustment. FFHQH includes 60,000 triplets for train, 5000 for validation and 5000 for test. Triplet stands for <real image, mask, composite image>.

Fingerprint inpainting and denoising

Synthetic training set: This set is constructed in the following two steps and will be used for estimation/training purposes. i) 84,000 275 pixel x 400 pixel ground-truth fingerprint images without any noise or scratches, but with random transformations (at most five pixels translation and +/-10 degrees rotation) were generated by using the software Anguli: Synthetic Fingerprint Generator. ii) 84,000 275 pixel x 400 pixel degraded fingerprint images were generated by applying random artifacts (blur, brightness, contrast, elastic transformation, occlusion, scratch, resolution, rotation) and backgrounds to the ground-truth fingerprint images. In total, it contains 168,000 fingerprint images (84,000 fingerprints, and two impressions - one ground-truth and one degraded - per fingerprint). Synthetic test set: This set is constructed similarly to the synthetic training set and will be used to evaluate the reconstruction performance. In total, it contains 16,800 fingerprint images (8,400 fingerprints and two impressions - one ground-truth and one degraded - per fingerprint). Since this set will be used for the purpose of evaluating the reconstruction performance, only the degraded and not the ground-truth fingerprint images will be provided to participants. Real test set: This set is constructed by systematically drawing fingerprint images with varying sizes from publicly available datasets. In total, it contains 1680 fingerprint images (140 fingerprints and 12 impressions - high-quality scans under operational conditions - per fingerprint). Description from: Fingerprint inpainting and denoising (WCCI'18, ECCV'18)

FLUXSynID

The FLUXSynID Dataset comprises 14,889 high-resolution synthetic face identities, each uniquely represented with paired images: document-style images and trusted live-capture images. The document-style images are designed to simulate ID photos, while the live-capture images are generated to reflect real-world conditions, including variations in lighting, facial expressions, and camera angles. This dual representation facilitates robust benchmarking for face recognition, morphing attack detection, and cross-domain identity verification tasks. FLUXSynID leverages cutting-edge generative models to ensure high fidelity and diversity across identities, enabling comprehensive analysis of biometric systems under varied conditions. For more information and to generate your own custom datasets using the FLUXSynID framework, visit the FLUXSynID GitHub repository.

geometricShapes14k

The dataset was curated from the 1% data sample file of the Wikipedia-based Image Text (WIT) Dataset. Captions for images were generated using the BLIP model. Control images were derived using the Primitive software, resulting in a dataset of 14,279 pairs of control and target images with captions.

GLAMI-1M

We introduce GLAMI-1M: the largest multilingual image-text classification dataset and benchmark. The dataset contains images of fashion products with item descriptions, each in 1 of 13 languages. Categorization into 191 classes has high-quality annotations: all 100k images in the test set and 75% of the 1M training set were human-labeled. The paper presents baselines for image-text classification showing that the dataset presents a challenging fine-grained classification problem: The best scoring EmbraceNet model using both visual and textual features achieves 69.7% accuracy. Experiments with a modified Imagen model show the dataset is also suitable for image generation conditioned on text.

HRI

The HRI Dataset comprises a total of 3,200 image pairs. Each image pair comprises a clean background image, a depth image, a rain layer mask image, and a rainy image. It contains three scenes: lane, citystreet and japanesestreet, with image resolutions of $2048\times1024$. The lane scene contains 1,600 image pairs, consisting of images from 4 camera viewpoints, each viewpoint containing 100 images of different moments, and each moment containing 4 different intensities of rainy scenes. The citystreet scene contains 600 image pairs, consisting of images from 6 camera viewpoints, each viewpoint containing 25 images of different moments, and each moment containing 4 different intensities of rainy scenes. The japanesestreet scene contains 1,000 image pairs, consisting of images from 10 camera viewpoints, each viewpoint containing 25 images of different moments, and each moment containing 4 different intensities of rainy scenes. We split the HRI dataset into training set and test set according to camera viewpoints. For the lane scene, the training set contains images from 3 camera viewpoints, and the test set contains images from 1 camera viewpoint. For the citystreet scene, the training set contains images from 5 camera viewpoints, and the test set contains images from 1 camera viewpoint. For the japanesestreet scene, the training set contains images from 8 camera viewpoints, and the test set contains images from 2 camera viewpoints. Therefore, the training set comprises a total of 2,500 image pairs, and the test set comprises a total of 700 image pairs.

HumanRig

Overview - Dataset Name: HumanRig - Paper: CVPR2025 - "HumanRig: Learning Automatic Rigging for Humanoid Character in a Large Scale Dataset" - Authors: [Zedong Chu · Feng Xiong · Meiduo Liu · Jinzhi Zhang · Mingqi Shao · Zhaoxu Sun · Di Wang · Mu Xu] - The HumanRig dataset is introduced in the CVPR2025 paper titled "HumanRig: Learning Automatic Rigging for Humanoid Character in a Large Scale Dataset". This work addresses the critical need for a comprehensive dataset and a robust framework for automatic rigging of 3D humanoid character models. HumanRig is a large-scale dataset of AI-generated T-pose humanoid models, all rigged with a consistent skeleton topology. It significantly surpasses previous datasets in terms of size, diversity, complexity, and practical motion applications. ### Dataset Details - Number of Samples: 11,434 - Data Splits: - Training Set: 0-9147 (80%) - Validation Set: 9148-10290 (10%) - Test Set: 10291-11433 (10%) ### Data Structure Each annotated sample in the HumanRig dataset includes the following components: - Rigged T-pose Humanoid Mesh: A rigged 3D mesh of the humanoid character in T-pose. - 3D Skeleton Joint Positions: The positions of the joints in the 3D skeleton. - Skinning Weight Matrix: A matrix that defines the influence of each joint on the vertices of the mesh. - Front-view Image: A 2D image of the humanoid character from the front view. - Camera Parameters: The parameters of the camera used to capture the front-view image. - 2D Skeleton Joint Positions: The positions of the joints in the 2D front-view image. ### Skeleton Joint Order The skeleton joints in the dataset follow a specific order: python ['Hips', 'Spine', 'Spine1', 'Spine2', 'Neck', 'Head', 'LeftShoulder', 'LeftArm', 'LeftForeArm', 'LeftHand', 'RightShoulder', 'RightArm', 'RightForeArm', 'RightHand', 'LeftUpLeg', 'LeftLeg', 'LeftFoot', 'LeftToeBase', 'RightUpLeg', 'RightLeg', 'RightFoot', 'RightToeBase'] ### Citation If you use the HumanRig dataset in your research, please cite the following paper: bibtex @article{chu2024humanrig, title={HumanRig: Learning Automatic Rigging for Humanoid Character in a Large Scale Dataset}, author={Chu, Zedong and Xiong, Feng and Liu, Meiduo and Zhang, Jinzhi and Shao, Mingqi and Sun, Zhaoxu and Wang, Di and Xu, Mu}, journal={arXiv preprint arXiv:2412.02317}, year={2024} }

iDesigner

Fashion trends are constantly evolving, but a trained eye can estimate with some accuracy the signature elements of a particular designer's style. 50,000 runway images, spanning 50 fashion designers, from our large repository of proprietary front row images from fashion shows over the past 15 years. The data includes a variety of fashion items, including: shoes, bags, dress, jackets etc.

InstructPix2Pix Image Editing Dataset

Dataset Figure A dataset for image editing containing >450k samples of: 1. input image (with corresponding text caption describing the image) 2. text-based edit instruction 3. edited image (with corresponding text caption describing the image) This dataset is automatically generated using a combination of GPT-3 (for generating the text edits) and StableDiffusion+Prompt-To-Prompt (for generating the input & edited images). Full description of the dataset can be found in the paper: (https://www.timothybrooks.com/instruct-pix2pix/)

Large Labelled Logo Dataset (L3D)

It is composed of around 770k of color 256x256 RGB images extracted from the European Union Intellectual Property Office (EUIPO) open registry. Each of them is associated to multiple labels that classify the figurative and textual elements that appear in the images. These annotations have been classified by the EUIPO evaluators using the Vienna classification, a hierarchical classification of figurative marks. We suggest it to be used for: 1. Unconditional trademark generation 2. Conditional trademark generation. 3. Multi-label logo classification (Vienna classification). 4. Optical Character Recognition. 5. Conditional trade 6. Image segmentation. 7. Image retrieval

LISA Gaze Dataset

LISA Gaze is a dataset for driver gaze estimation comprising of 11 long drives, driven by 10 subjects in two different cars. Source: Driver Gaze Zone Estimation using Convolutional Neural Networks: A General Framework and Ablative Analysis

LLMafia

To evaluate our proposed strategy of asynchronous communication for LLMs, we run games of Mafia with human players, incorporating an LLM-based agent as an additional player, within an asynchronous chat environment. The data includes all players messages and votes including timestamps, game management messages (e.g., announcements of the beginning and end of phases), in addition to records related to the agent, such as the prompts that were provided at each timestamp. While prior datasets for Mafia exist, ours is the first to integrate an LLM agent, allowing analysis of human-LLM interactions. Detailed analysis of the data is available in our paper: Time to Talk: LLM Agents for Asynchronous Group Communication in Mafia Games

MeshFLeet

MeshFleet is a filtered and annotated dataset of High Quality vehicles derived from Objaverse XL. It contains the sha256 of the objects together with consitent object captions and vehicle parameters.

MFSD

During the covid-19 era wearing face masks posed new challenges to face-related tasks, including facial recognition, face inpainting, expression recognition, and object removal. Mask region segmentation is a preliminary stage to tackle the occlusion issue corresponding to the face-related tasks. Existing masked face datasets are not procedure binary segmentation maps because Segmenting mask regions manually is a time-consuming operation. As a result, existing unmasking methods; synthesize training data by overlaying masks on existing face datasets. However, since these techniques rely on an artificially generated mask, their effects tend to seem unnatural. To address this issue, the masked face segmentation dataset(MFSD) provides the first public training dataset for the mask segmentation task.

MH-FED

This dataset provides a collection of 162K images and 70 Videos of Meta-Humans. There are 10 Highly realistic Meta-Humans expressing 7 facial expressions.

Mpm-Verse-Large

This dataset contains Material-Point-Method (MPM) simulations for various materials, including water, sand, plasticine, jelly, and rigid collisions. Each material is represented as point-clouds that evolve over time. The dataset is designed for learning and predicting MPM-based physical simulations. Each material contains 50 trajectories with different initial velocity field.

출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.