(ECCV 2020) Conditional Sequential Modulation for Efficient Global Image Retouching
-
Updated
Sep 28, 2020 - Python
(ECCV 2020) Conditional Sequential Modulation for Efficient Global Image Retouching
Implementation of the CSRNet paper (CVPR 18) in keras-tensorflow
SOFT-CSRNET : Counting people in drone video footage
An unofficial implementation of CSRNet for crowd counting.
Real-time crowd density estimation & interactive flow simulation powered by CSRNet. Trained on ShanghaiTech and custom Indian crowd datasets with FastAPI & Vite.
Train a CSRNet for estimating a crowd distribution
Implement CSRnet for counting vehicle traffic
A deep learning project implementing CSRNet for accurate people counting in highly congested scenes, based on the architecture described in the CSRNet paper.
Keras based implementation of the CSRNET model
Real-time crowd intelligence platform — YOLOv8 + CSRNet + multi-horizon LSTM risk forecasting
AI-powered smart city crowd management and dynamic bus dispatch system using CSRNet density map regression to monitor transit stop overcrowding and optimize public transportation.
Crowd counting and behavior analysis web app using PyTorch (CSRNet) with a Flask-based frontend.
Dense crowd counting with CSRNet + CBAM attention — MAE 61.14 / RMSE 90.91 on ShanghaiTech Part A, beating the CVPR'18 CSRNet baseline (68.2 / 115.0) — paired with a fully local Ollama multimodal agent that explains every prediction.
Deep learning-based crowd counting using a VGG16 frontend and dilated convolutions. Supports image and video analysis, density heatmap visualization, frame-by-frame counting, charts, and CSV export using Streamlit, PyTorch, and OpenCV.
Address the crowd counting problem on the Mall dataset (sparse) by exploring regression-based (Xception) and density-based (CSRNet) approaches.
A Django web application providing real-time crowd counting and density heatmap overlay streaming using pre-trained CSRNet models.
PyTorch training and evaluation framework for crowd counting density estimation models (CSRNet, resSFCN, MCNN) on standard and custom datasets.
Single-Image Crowd Counting - Computer Vision Project
Reproducible PyTorch implementation of CSRNet for crowd counting and density map estimation
To associate your repository with the csrnet topic, visit your repo's landing page and select "manage topics."