Research map / Video / Video segmentation and tracking
Understanding Traffic Density From Large-Scale Web Camera Data
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use FCN-MT instead of Baseline 2 (Hydra) | Traffic density estimation · TRANCOS | MAE: 10.99 -> 5.31 | Helped |
| Jointly learn vehicle density and global count rather than density alone (ablation) | FCN-MT · TRANCOS | MAE: 5.47 -> 5.31 | Helped |
| Use FCN-MT instead of Baseline 2 (Hydra) | Traffic density estimation · WebCamT Downtown | MAE: 3.55 -> 2.74 | Helped |
| Use FCN-MT instead of Baseline 2 (Hydra) | Traffic density estimation · WebCamT Parkway | MAE: 3.64 -> 2.52 | Helped |
| Use FCN-MT instead of Cross-scene DNN | Crowd counting · UCSD | MAE: 1.6 -> 1.67 | Hurt |
| Use FCN-MT instead of Cross-scene DNN | Crowd counting · UCSD | MSE: 3.31 -> 3.41 | Hurt |
| Use block-specific weights with a rank constraint instead of uniform image-wide weights | OPT-RC · WebCamT | MAE: 5.91 -> 4.56; 5.19 -> 4.24 | Helped |
| Use FCN-MT instead of OPT-RC | Traffic density estimation · TRANCOS | MAE: 12.41 -> 5.31 | Helped |
| Use FCN-MT instead of OPT-RC | Crowd counting · UCSD | MAE: 2.03 -> 1.67 | Helped |
Closest papers by meaning
- STREETS: A Novel Camera Network Dataset for Traffic Flow (NeurIPS 2019 · similarity 0.86)
- Highway Vehicle Counting in Compressed Domain (CVPR 2016 · similarity 0.85)
- CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenes (CVPR 2018 · similarity 0.84)
- Traffic Scene Parsing through the TSP6K Dataset (CVPR 2024 · similarity 0.84)
- TUMTraf VideoQA: Dataset and Benchmark for Unified Spatio-Temporal Video Understanding in Traffic Scenes (ICML 2025 · similarity 0.83)
- On-Road Object Importance Estimation: A New Dataset and A Model with Multi-Fold Top-Down Guidance (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.