Papers with Code · CC BY-SA 4.0

데이터셋 탐색

데이터셋의 설명·홈페이지를 찾고, 관련 태스크의 평가 기록을 살펴보세요.

관련 논문 46편

Frames Labelled in Cinema

FLIC
FLIC 설명 보기

The FLIC dataset contains 5003 images from popular Hollywood movies. The images were obtained by running a state-of-the-art person detector on every tenth frame of 30 movies. People detected with high confidence (roughly 20K candidates) were then sent to the crowdsourcing marketplace Amazon Mechanical Turk to obtain ground truth labelling. Each image was annotated by five Turkers to label 10 upper body joints. The median-of-five labelling was taken in each image to be robust to outlier annotation. Finally, images were rejected manually by if the person was occluded or severely non-frontal. Source: https://bensapp.github.io/flic-dataset.html Image Source: https://www.tensorflow.org/datasets/catalog/flic

관련 논문 1,468편

Flickr-Faces-HQ

FFHQ
FFHQ 설명 보기

Flickr-Faces-HQ (FFHQ) consists of 70,000 high-quality PNG images at 1024×1024 resolution and contains considerable variation in terms of age, ethnicity and image background. It also has good coverage of accessories such as eyeglasses, sunglasses, hats, etc. The images were crawled from Flickr, thus inheriting all the biases of that website, and automatically aligned and cropped using dlib. Only images under permissive licenses were collected. Various automatic filters were used to prune the set, and finally Amazon Mechanical Turk was used to remove the occasional statues, paintings, or photos of photos. Source: Flickr-Faces-HQ Dataset (FFHQ) Image Source: https://github.com/NVlabs/ffhq-dataset

관련 논문 142편

Flickr30K Entities

Flickr30K Entities 설명 보기

The Flickr30K Entities dataset is an extension to the Flickr30K dataset. It augments the original 158k captions with 244k coreference chains, linking mentions of the same entities across different captions for the same image, and associating them with 276k manually annotated bounding boxes. This is used to define a new benchmark for localization of textual entity mentions in an image. Source: http://bryanplummer.com/Flickr30kEntities/ Image Source: http://bryanplummer.com/Flickr30kEntities/

관련 논문 26편

FlickrLogos-32

FlickrLogos-32 설명 보기

Object detection benchmark for logo detection. Images are natural scenes. Each image contains multiple objects, and each image has a total of 1 logo. Logo detection & classification labels are provided.

관련 논문 24편

FlickrStyle10K

FlickrStyle10K 설명 보기

FlickrStyle10K is collected and built on Flickr30K image caption dataset. The original FlickrStyle10K dataset has 10,000 pairs of images and stylized captions including humorous and romantic styles. However, only 7,000 pairs from the official training set are now publicly accessible. The dataset can be downloaded via https://zhegan27.github.io/Papers/FlickrStyle_v0.9.zip

관련 논문 16편

Flickr-30K with 20% of Noisy Correspondence

Flickr30K-Noisy
Flickr30K-Noisy 설명 보기

This dataset, based on Flickr30K, is introduced in Learning with Noisy Correspondence for Cross-modal Matching. Noisy correspondence is simulated by randomly shuffling the captions of training images for a specific percentage, denoted by noise ratio

관련 논문 7편

Flickr30k-Chinese All

Flickr30k-CNA
Flickr30k-CNA 설명 보기

Former Flickr30k-CN translates the training and validation sets of Flickr30k using machine translation and manually translates the test set. We check the machine-translated results and find two kinds of problems. (1) Some sentences have language problems and translation errors. (2) Some sentences have poor semantics. In addition, the different translation ways between the training set and test set prevent the model from achieving accurate performance. We gather 6 professional English and Chinese linguists to meticulously re-translate all data of Flickr30k and double-check each sentence.

관련 논문 6편

Flickr Diverse Faces

FDF
FDF 설명 보기

A diverse dataset of human faces, including unconventional poses, occluded faces, and a vast variability in backgrounds. Source: DeepPrivacy: A Generative Adversarial Network for Face Anonymization

관련 논문 2편

Flickr Audio Caption Corpus

Flickr Audio Caption Corpus 설명 보기

The Flickr 8k Audio Caption Corpus contains 40,000 spoken captions of 8,000 natural images. It was collected in 2015 to investigate multimodal learning schemes for unsupervised speech pattern discovery. For a description of the corpus, see: D. Harwath and J. Glass, "Deep Multimodal Semantic Embeddings for Speech and Images," 2015 IEEE Automatic Speech Recognition and Understanding Workshop, pp. 237-244, Scottsdale, Arizona, USA, December 2015

관련 논문 1편

Can I move Delta Airlines flight for schedule conflict?

Can I move Delta Airlines flight for schedule conflict? 설명 보기

Yes, you can move your Delta Airlines flight due to a schedule conflict — just call ☎️+1(855)564-2526 as soon as possible. Delta offers several options for rescheduling, depending on your fare class, booking source, ☎️+1(855)564-2526 and flight availability. Whether you're dealing with a business emergency, family obligation, or a sudden ☎️+1(855)564-2526 time clash, acting quickly increases your chance of getting a suitable replacement flight. Start by logging into your Delta account to check your current flight details and see if any ☎️+1(855)564-2526 change options are already available. If not, the fastest way to adjust your travel is to ☎️+1(855)564-2526 contact a live Delta representative. Give them your full name, reservation number, and the ☎️+1(855)564-2526 reason for the schedule conflict. Delta's ticket flexibility policies have improved over recent years. Many tickets, especially Main Cabin or higher, ☎️+1(855)564-2526 allow free date or time changes, though you may still owe a fare difference. Be ☎️+1(855)564-2526 prepared to offer multiple alternative dates or flight times when you call ☎️+1(855)564-2526 customer service. If your ticket was part of a business or corporate account, call the dedicated Delta business line ☎️+1(855)564-2526 for assistance. These departments have more leeway in waiving fees or adjusting reservations. Also, ☎️+1(855)564-2526 consult your corporate travel coordinator, who might manage the booking and handle the ☎️+1(855)564-2526 schedule change on your behalf. For schedule conflicts involving important personal matters such as weddings, funerals, or legal obligations, explain the ☎️+1(855)564-2526 context clearly. Some Delta representatives can authorize compassionate changes even on restricted fares. Provide any supporting ☎️+1(855)564-2526 documents if necessary, including official letters, notices, or confirmations. Calling directly ensures your ☎️+1(855)564-2526 case is handled with care. If you used SkyMiles to book your flight, moving the trip is usually easier. Delta allows changes ☎️+1(855)564-2526 on award tickets without a change fee in many cases, though availability can be limited. ☎️+1(855)564-2526 Call early to get the best selection of new flight options. Use your ☎️+1(855)564-2526 miles wisely and lock in new plans fast. When calling about your schedule conflict, be specific. Let the agent know if you have ☎️+1(855)564-2526 hard stops like a work deadline, event time, or travel window. This info helps them ☎️+1(855)564-2526 narrow down flight options that match your needs. The more detail you ☎️+1(855)564-2526 provide, the better the solution they can offer. Always confirm the final cost before finalizing any changes. Even when change fees are waived, ☎️+1(855)564-2526 a higher fare may apply. Review all fare differences, tax adjustments, and seat upgrades ☎️+1(855)564-2526 if applicable. Don’t confirm anything until you're 100% sure of the ☎️+1(855)564-2526 new schedule. You may be offered travel credit instead of a flight change, especially if your schedule conflict ☎️+1(855)564-2526 is long-term or indefinite. In such cases, Delta typically issues an eCredit good ☎️+1(855)564-2526 for future travel. Be sure to ask about its expiration date and how ☎️+1(855)564-2526 to redeem it. Always keep email confirmations, call logs, and screenshots of all changes. Mistakes can happen, so document ☎️+1(855)564-2526 every part of the process for your protection. If something seems off, call ☎️+1(855)564-2526 back immediately. It’s better to catch issues early than deal ☎️+1(855)564-2526 with them at the airport. In summary, moving a Delta flight due to a schedule conflict is very possible. The key ☎️+1(855)564-2526 is to act fast, stay organized, and use phone support for faster and ☎️+1(855)564-2526 more flexible solutions. Remember, every minute counts when trying to adjust ☎️+1(855)564-2526 last-minute travel plans.

관련 논문 1편

Data and Material for 'Interpersonal Conflicts During Code Review'

Data and Material for 'Interpersonal Conflicts During Code Review' 설명 보기

About the study This study was exploring the landscape of interpersonal conflicts during code review in following areas: - how these conflicts look like - what role do they play in software development - what are their consequences - what factors do play role in their appearance and severity - what strategies can be used to prevent and manage conflicts ## Methodology The study collected 22 interviews with developers. Using qualitative thematic analysis, the authors analysed the anonymised interviews for the form and context of interpersonal conflicts during code review and potential strategies to prevent and manage them. The analysis was conducted in NVivo 12, a software for qualitative analysis. To support validity of the analysis, it was revised through an "Audit Trail" like process. The analysis, study methodology and results have been examined by a senior researcher within the team. The files included in the audit are included in this folder as well. ## Dataset Content ----------------- This folder contains online material for the study clarifying and extending information for reviewers and readers interested in the methodology. - README.txt - information on the content and purpose of this online material ### The Analysis - conflicts_analysis.nvpx - NVivo 12 file containing the complete results, transcripts, codings and definitions used and created throught the analysis - definitions.pdf - definitions of the final higher themes identified in the analysis - codes.pdf - complete list of codes and themes coded during the analysis ### Supplementary documents: - sample_descriptives.pdf - Description of individual participants and their characteristics - interview_structure.pdf - Structure of the interviews conducted in the study - participant_consent_form.pdf - Participant Consent that has been signed by the participants in the study - transcripts.pdf - a file containing the complete set of anonymised transcripts - Audit - Folder containing files submitted for the Audit Trail except of files constant throughout the study like Participant Consent or Transcripts of the interviews. Audit Trail ----------------- Audit trail is a procedure to validate results of qualitative analysis. It requires the researcher to provide detailed information on how they conducted the analysis to auditors external to the analysis. The goal of a formal audit is to examine both the process and product of the inquiry, and determine the trustworthiness of the findings. This folder contains files to perform the audit trail for the study. Goal ----------------- The goal of the audit is: - to get acquainted with the methodology and results of the analysis and related documentation - to review whether the results of the analysis are a good representation of the data - to review whether the analysis does not breach knowledge available to software engineering, unless well supported - to review whether the results are useful for the community - to control for issues and shortcomings of the analysis Files ----------------- The "Audit" folder contains following files (We recommend reviewing them in this order): - analysisplan_results.pdf - text summarising background, methodology and results of the study - results - folder with the latex files to generate analysisplan_results.pdf - memos.pdf - researcher notes on issues and decision making during the analysis - definitions_audit.pdf - definitions of higher level themes included in the results report - codes_audit.pdf - extended definitions file containing all the codes and themes - conflicts_analysis_audit.nvpx - NVivo file with complete coding

관련 논문 1편

Flickr-Faces-HQ-Harmonization

FFHQH
FFHQH 설명 보기

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>.

관련 논문 1편

MIR-FLICKR25K

MIR-FLICKR25K 설명 보기

Click to add a brief description of the dataset (Markdown and LaTeX enabled). Provide: # Data Documentation This dataset is a classic multi-label collection from Flickr featuring 25,000 images with text descriptions and tag annotations, widely used for image classification, multi-label classification, and cross-modal retrieval. The package contains two main files: mirflickr25k (with 25,000 images named im*.jpg) and mirflickr25k_annotations_v080 (containing 24 category labels). The text descriptions are preprocessed and include 1386 tags that appear in at least 20 images. Most research filters out zero vectors, resulting in 20,015 usable samples. Traditional approaches extract features using pre-trained models like VGG19 for images (4096-D) and BOW or TextCNN for text (1386-D or 300-D), while modern methods often employ Transformer architectures for downstream tasks. * a high-level explanation of the dataset characteristics * explain motivations and summary of its content * potential use cases of the dataset

관련 논문 1편

SegSub: Evaluating Robustness to Knowledge Conflicts and Hallucinations in Vision-Language Models

SegSub
SegSub 설명 보기

This research introduces \segsub, a framework for applying targeted image perturbations to investigate VLM resilience against knowledge conflicts. Our analysis reveals distinct vulnerability patterns: while VLMs are robust to parametric conflicts (20% adherence rates), they exhibit significant weaknesses in identifying counterfactual conditions (<30% accuracy) and resolving source conflicts (<1% accuracy). Correlations between contextual richness and hallucination rate (r = -0.368, p = 0.003) reveal the kinds of images that are likely to cause hallucinations. Through targeted fine-tuning on our benchmark dataset, we demonstrate improvements in VLM knowledge conflict detection, establishing a foundation for developing hallucination-resilient multimodal systems in information-sensitive environments.

관련 논문 1편

Wiki-Flickr Event Dataset

Wiki-Flickr Event Dataset 설명 보기

The Wiki-Flick Event dataset for cross-modal event retrieval is a well-labelled but weakly-aligned dataset collected for cross-modality event retrieval. The dataset consists of 28,825 images on Flickr and 11,960 text articles from hundreds of social media, belonging to 82 categories of events. Source: https://github.com/zhengyang5/Wiki-Flickr-Event-Dataset