{"url":"/dataset/shanghaitech","name":"ShanghaiTech","full_name":null,"description_markdown":"The Shanghaitech dataset is a large-scale crowd counting dataset. It consists of 1198 annotated crowd images. The dataset is divided into two parts, Part-A containing 482 images and Part-B containing 716 images. Part-A is split into train and test subsets consisting of 300 and 182 images, respectively. Part-B is split into train and test subsets consisting of 400 and 316 images. Each person in a crowd image is annotated with one point close to the center of the head. In total, the dataset consists of 330,165 annotated people. Images from Part-A were collected from the Internet, while images from Part-B were collected on the busy streets of Shanghai.\r\n\r\nSource: [Iterative Crowd Counting](https://arxiv.org/abs/1807.09959)\r\n\r\nImage Source: [Li et al](https://www.researchgate.net/figure/Test-images-from-the-ShanghaiTech-A-28-dataset-The-goal-of-this-paper-is-to-calculate_fig1_322652466)","description_withheld":null,"homepage":"https://github.com/desenzhou/ShanghaiTechDataset","introduced_date":"2016-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/single-image-crowd-counting-via-multi-column-1","title":"Single-Image Crowd Counting via Multi-Column Convolutional Neural Network","first_author":"Yingying Zhang","url":null},"license":{"name":"Unknown","url":null},"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[{"name":"Anomaly Detection","url":"/task/anomaly-detection","datasets_with_task":"/datasets/task/anomaly-detection"},{"name":"Video Anomaly Detection","url":"/task/video-anomaly-detection","datasets_with_task":"/datasets/task/video-anomaly-detection"},{"name":"Crowd Counting","url":"/task/crowd-counting","datasets_with_task":"/datasets/task/crowd-counting"},{"name":"Anomaly Detection In Surveillance Videos","url":"/task/anomaly-detection-in-surveillance-videos","datasets_with_task":"/datasets/task/anomaly-detection-in-surveillance-videos"},{"name":"Abnormal Event Detection In Video","url":"/task/abnormal-event-detection-in-video","datasets_with_task":"/datasets/task/abnormal-event-detection-in-video"},{"name":"Cross-Part Crowd Counting","url":"/task/cross-part-crowd-counting","datasets_with_task":"/datasets/task/cross-part-crowd-counting"}],"languages":[],"variants":["ShanghaiTech","ShanghaiTech A","ShanghaiTech B"],"data_loaders":[{"repo":"https://github.com/desenzhou/ShanghaiTechDataset","url":"https://github.com/desenzhou/ShanghaiTechDataset","frameworks":[]}],"num_papers_in_archive":277,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/crowd-counting-on-shanghaitech-a","task":"Crowd Counting","dataset_variant":"ShanghaiTech A","rows":35,"metrics":["MAE","MSE","RMSE"],"first_row_in_archive_order":{"model":"EBC-ZIP-B","paper":"/paper/ebc-zip-improving-blockwise-crowd-counting","metrics":{"MAE":"47.81","RMSE":"75.04"},"code_links":[{"title":"yiming-m/ebc-zip","url":"https://github.com/yiming-m/ebc-zip"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/crowd-counting-on-shanghaitech-b","task":"Crowd Counting","dataset_variant":"ShanghaiTech B","rows":32,"metrics":["MAE","MSE","RMSE"],"first_row_in_archive_order":{"model":"EBC-ZIP-B","paper":"/paper/ebc-zip-improving-blockwise-crowd-counting","metrics":{"MAE":"5.51","RMSE":"8.63"},"code_links":[{"title":"yiming-m/ebc-zip","url":"https://github.com/yiming-m/ebc-zip"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/anomaly-detection-on-shanghaitech","task":"Anomaly Detection","dataset_variant":"ShanghaiTech","rows":31,"metrics":["AUC","RBDC","TBDC"],"first_row_in_archive_order":{"model":"DAC(STG-NF + Jigsaw)","paper":"/paper/divide-and-conquer-in-video-anomaly-detection","metrics":{"AUC":"87.72%"},"code_links":[{"title":"XiaoJian923/Divide-and-Conquer","url":"https://github.com/XiaoJian923/Divide-and-Conquer"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/video-anomaly-detection-on-shanghaitech-4","task":"Video Anomaly Detection","dataset_variant":"ShanghaiTech","rows":7,"metrics":["AUC","RBDC","TBDC"],"first_row_in_archive_order":{"model":"MULDE-object-centric-micro","paper":"/paper/mulde-multiscale-log-density-estimation-via","metrics":{"AUC":"86.7%"},"code_links":[{"title":"jakubmicorek/MULDE-Multiscale-Log-Density-Estimation-via-Denoising-Score-Matching-for-Video-Anomaly-Detection","url":"https://github.com/jakubmicorek/MULDE-Multiscale-Log-Density-Estimation-via-Denoising-Score-Matching-for-Video-Anomaly-Detection"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/anomaly-detection-in-surveillance-videos-on-8","task":"Anomaly Detection In Surveillance Videos","dataset_variant":"ShanghaiTech","rows":1,"metrics":["AUC"],"first_row_in_archive_order":{"model":"LMM_VAD","paper":"/paper/10-security-and-privacy-problems-in-self","metrics":{"AUC":"88.8"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/ebc-zip-improving-blockwise-crowd-counting","title":"EBC-ZIP: Improving Blockwise Crowd Counting with Zero-Inflated Poisson Regression","date":"2025-06-24","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/anyanomaly-zero-shot-customizable-video-1","title":"AnyAnomaly: Zero-Shot Customizable Video Anomaly Detection with LVLM","date":"2025-03-06","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/videopatchcore-an-effective-method-to","title":"VideoPatchCore: An Effective Method to Memorize Normality for Video Anomaly Detection","date":"2024-09-24","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/follow-the-rules-reasoning-for-video-anomaly","title":"Follow the Rules: Reasoning for Video Anomaly Detection with Large Language Models","date":"2024-07-14","rows_on_this_dataset":2,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":2,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/bounding-boxes-and-probabilistic-graphical","title":"Bounding Boxes and Probabilistic Graphical Models: Video Anomaly Detection Simplified","date":"2024-07-08","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/improving-point-based-crowd-counting-and","title":"Improving Point-based Crowd Counting and Localization Based on Auxiliary Point Guidance","date":"2024-05-17","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/crowd-counting-and-individual-localization","title":"Crowd Counting and Individual Localization Using Pseudo Square Label","date":"2024-05-13","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/vmambacc-a-visual-state-space-model-for-crowd","title":"VMambaCC: A Visual State Space Model for Crowd Counting","date":"2024-05-07","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/an-exploratory-study-on-human-centric-video","title":"An Exploratory Study on Human-Centric Video Anomaly Detection through Variational Autoencoders and Trajectory Prediction","date":"2024-04-29","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/mulde-multiscale-log-density-estimation-via","title":"MULDE: Multiscale Log-Density Estimation via Denoising Score Matching for Video Anomaly Detection","date":"2024-03-21","rows_on_this_dataset":4,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/clip-ebc-clip-can-count-accurately-through","title":"CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification","date":"2024-03-14","rows_on_this_dataset":9,"code_links":1,"syntology":null},{"paper":"/paper/making-anomalies-more-anomalous-video-anomaly","title":"Making Anomalies More Anomalous: Video Anomaly Detection Using a Novel Generator and Destroyer","date":"2024-02-26","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/fgenet-fine-grained-extraction-network-for-2","title":"FGENet: Fine-Grained Extraction Network for Congested Crowd Counting","date":"2024-01-02","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/divide-and-conquer-in-video-anomaly-detection","title":"Divide and Conquer in Video Anomaly Detection: A Comprehensive Review and New Approach","date":"2023-09-26","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/diversity-measurable-anomaly-detection","title":"Diversity-Measurable Anomaly Detection","date":"2023-03-09","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":11,"samples_ran":6,"samples_unverified":5,"pointer_only_for_licence":11,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/improving-deep-regression-with-ordinal","title":"Improving Deep Regression with Ordinal Entropy","date":"2023-01-21","rows_on_this_dataset":2,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":5,"samples_ran":4,"samples_unverified":1,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/eval-explainable-video-anomaly-localization","title":"EVAL: Explainable Video Anomaly Localization","date":"2022-12-15","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/attribute-based-representations-for-accurate","title":"An Attribute-based Method for Video Anomaly Detection","date":"2022-12-01","rows_on_this_dataset":1,"code_links":4,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":2,"samples_unverified":1,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/normalizing-flows-for-human-pose-anomaly","title":"Normalizing Flows for Human Pose Anomaly Detection","date":"2022-11-20","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/spatio-temporal-predictive-tasks-for-abnormal","title":"Spatio-temporal predictive tasks for abnormal event detection in videos","date":"2022-10-27","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/improving-local-features-with-relevant","title":"Improving Local Features with Relevant Spatial Information by Vision Transformer for Crowd Counting","date":"2022-09-30","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/self-supervised-masked-convolutional","title":"Self-Supervised Masked Convolutional Transformer Block for Anomaly Detection","date":"2022-09-25","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/context-recovery-and-knowledge-retrieval-a","title":"Context Recovery and Knowledge Retrieval: A Novel Two-Stream Framework for Video Anomaly Detection","date":"2022-09-07","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/video-anomaly-detection-by-solving-decoupled","title":"Video Anomaly Detection by Solving Decoupled Spatio-Temporal Jigsaw Puzzles","date":"2022-07-20","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/ssmtl-revisiting-self-supervised-multi-task","title":"SSMTL++: Revisiting Self-Supervised Multi-Task Learning for Video Anomaly Detection","date":"2022-07-16","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/rethinking-spatial-invariance-of-1","title":"Rethinking Spatial Invariance of Convolutional Networks for Object Counting","date":"2022-06-10","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/attention-based-residual-autoencoder-for","title":"Attention-based residual autoencoder for video anomaly detection","date":"2022-05-25","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/fusioncount-efficient-crowd-counting-via","title":"FusionCount: Efficient Crowd Counting via Multiscale Feature Fusion","date":"2022-02-28","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/iterative-correlation-based-feature","title":"Few-shot Object Counting with Similarity-Aware Feature Enhancement","date":"2022-01-22","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/regularity-learning-via-explicit-distribution","title":"Regularity Learning via Explicit Distribution Modeling for Skeletal Video Anomaly Detection","date":"2021-12-07","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/self-supervised-predictive-convolutional","title":"Self-Supervised Predictive Convolutional Attentive Block for Anomaly Detection","date":"2021-11-17","rows_on_this_dataset":2,"code_links":4,"syntology":null},{"paper":"/paper/ubnormal-new-benchmark-for-supervised-open","title":"UBnormal: New Benchmark for Supervised Open-Set Video Anomaly Detection","date":"2021-11-16","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/10-security-and-privacy-problems-in-self","title":"10 Security and Privacy Problems in Large Foundation Models","date":"2021-10-28","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/rethinking-counting-and-localization-in","title":"Rethinking Counting and Localization in Crowds:A Purely Point-Based Framework","date":"2021-07-27","rows_on_this_dataset":1,"code_links":3,"syntology":null},{"paper":"/paper/anomaly-detection-in-video-via-self","title":"Anomaly Detection in Video via Self-Supervised and Multi-Task Learning","date":"2020-11-15","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/distribution-matching-for-crowd-counting","title":"Distribution Matching for Crowd Counting","date":"2020-09-28","rows_on_this_dataset":2,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":5,"samples_ran":4,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/a-scene-agnostic-framework-with-adversarial","title":"A Background-Agnostic Framework with Adversarial Training for Abnormal Event Detection in Video","date":"2020-08-27","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":3,"samples_unverified":0,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/any-shot-sequential-anomaly-detection-in","title":"Any-Shot Sequential Anomaly Detection in Surveillance Videos","date":"2020-04-05","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/encoder-decoder-based-convolutional-neural","title":"Encoder-Decoder Based Convolutional Neural Networks with Multi-Scale-Aware Modules for Crowd Counting","date":"2020-03-12","rows_on_this_dataset":2,"code_links":2,"syntology":null},{"paper":"/paper/segmentation-guided-attention-network-for","title":"Crowd Counting via Segmentation Guided Attention Networks and Curriculum Loss","date":"2019-11-18","rows_on_this_dataset":4,"code_links":1,"syntology":null},{"paper":"/paper/learning-spatial-awareness-to-improve-crowd","title":"Learning Spatial Awareness to Improve Crowd Counting","date":"2019-09-16","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/from-open-set-to-closed-set-counting-objects","title":"From Open Set to Closed Set: Counting Objects by Spatial Divide-and-Conquer","date":"2019-08-15","rows_on_this_dataset":2,"code_links":5,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":8,"samples_ran":3,"samples_unverified":5,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-timescale-trajectory-prediction-for","title":"Multi-timescale Trajectory Prediction for Abnormal Human Activity Detection","date":"2019-08-12","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/locate-size-and-count-accurately-resolving","title":"Locate, Size and Count: Accurately Resolving People in Dense Crowds via Detection","date":"2019-06-18","rows_on_this_dataset":2,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":12,"samples_ran":1,"samples_unverified":11,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/learning-regularity-in-skeleton-trajectories","title":"Learning Regularity in Skeleton Trajectories for Anomaly Detection in Videos","date":"2019-03-08","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/object-centric-auto-encoders-and-dummy","title":"Object-centric Auto-encoders and Dummy Anomalies for Abnormal Event Detection in Video","date":"2018-12-11","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":21,"samples_ran":0,"samples_unverified":21,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/context-aware-crowd-counting","title":"Context-Aware Crowd Counting","date":"2018-11-26","rows_on_this_dataset":2,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":2,"samples_ran":1,"samples_unverified":1,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/scale-aggregation-network-for-accurate-and","title":"Scale Aggregation Network for Accurate and Efficient Crowd Counting","date":"2018-09-01","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/iterative-crowd-counting","title":"Iterative Crowd Counting","date":"2018-07-26","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/divide-and-grow-capturing-huge-diversity-in","title":"Divide and Grow: Capturing Huge Diversity in Crowd Images with Incrementally Growing CNN","date":"2018-07-26","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/crowd-counting-with-deep-negative-correlation","title":"Crowd Counting With Deep Negative Correlation Learning","date":"2018-06-01","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/crowd-counting-via-adversarial-cross-scale","title":"Crowd Counting via Adversarial Cross-Scale Consistency Pursuit","date":"2018-06-01","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/stan-spatio-temporal-adversarial-networks-for","title":"STAN: Spatio-Temporal Adversarial Networks for Abnormal Event Detection","date":"2018-04-23","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/leveraging-unlabeled-data-for-crowd-counting","title":"Leveraging Unlabeled Data for Crowd Counting by Learning to Rank","date":"2018-03-08","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/csrnet-dilated-convolutional-neural-networks","title":"CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes","date":"2018-02-27","rows_on_this_dataset":2,"code_links":11,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":15,"samples_ran":5,"samples_unverified":10,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/a-revisit-of-sparse-coding-based-anomaly","title":"A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework","date":"2017-10-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/generating-high-quality-crowd-density-maps","title":"Generating High-Quality Crowd Density Maps using Contextual Pyramid CNNs","date":"2017-08-02","rows_on_this_dataset":2,"code_links":0,"syntology":null},{"paper":"/paper/switching-convolutional-neural-network-for","title":"Switching Convolutional Neural Network for Crowd Counting","date":"2017-08-01","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/cnn-based-cascaded-multi-task-learning-of","title":"CNN-based Cascaded Multi-task Learning of High-level Prior and Density Estimation for Crowd Counting","date":"2017-07-30","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/single-image-crowd-counting-via-multi-column-1","title":"Single-Image Crowd Counting via Multi-Column Convolutional Neural Network","date":"2016-01-01","rows_on_this_dataset":2,"code_links":4,"syntology":null},{"paper":"/paper/cross-scene-crowd-counting-via-deep","title":"Cross-Scene Crowd Counting via Deep Convolutional Neural Networks","date":"2015-06-01","rows_on_this_dataset":2,"code_links":0,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":13,"samples_harvested":89,"samples_ran":33,"samples_unverified":56,"pointer_only_for_licence":32,"papers_with_no_sample_that_ran":1,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}