How shreelakshmigbhat/carvision-cnn-yolo works
This is a single-file, Colab-oriented car-detection workflow using Ultralytics YOLOv8. Its principal path prepares annotated car images, trains a detector, then runs prediction on an image and saves/displays results. The sampled code shows Drive mounting, annotation joins, YOLO label conversion and copying, plus image prediction; README claims model setup, training, and image/video inference. Training and video details are not present in the excerpts, so those capabilities are included without asserting their internal wiring.
Main entry points: car_detection_yolo_(1).py Image prediction
Components 15
Dataset preparation
- Load dataset
car_detection_yolo_(1).py - Join annotations
car_detection_yolo_(1).py - Convert labels
car_detection_yolo_(1).py - Arrange dataset
car_detection_yolo_(1).py
YOLO workflow
- Configure YOLO
car_detection_yolo_(1).py - Train detector
car_detection_yolo_(1).py - Trained detector
car_detection_yolo_(1).py
Detection results
- Image prediction
car_detection_yolo_(1).py - Video inference
car_detection_yolo_(1).py - Detection results
car_detection_yolo_(1).py - Saved predictions
Other components
- Notebook user
- Google Drive
- Car images and annotations
- Ultralytics YOLO
Connections 20
- Notebook user → Load dataset: runs notebook
- Load dataset → Google Drive: mounts Drive
- Google Drive → Car images and annotations: provides files
- Load dataset → Car images and annotations: reads data
- Load dataset → Join annotations: loads annotations
- Join annotations → Convert labels: supplies boxes
- Convert labels → Arrange dataset: writes labels
- Arrange dataset → Google Drive: copies dataset
- Configure YOLO → Ultralytics YOLO: uses library
- Configure YOLO → Train detector: configures training
- Train detector → Ultralytics YOLO: trains with
- Train detector → Trained detector: produces model
- Trained detector → Image prediction: used for prediction
- Trained detector → Video inference: used for prediction
- Image prediction → Ultralytics YOLO: calls predict
- Video inference → Ultralytics YOLO: runs inference
- Image prediction → Detection results: returns detections
- Video inference → Detection results: produces detections
- Image prediction → Saved predictions: saves predictions
- Notebook user → Image prediction: requests detection