ssd_keras and SSD-Tensorflow
These are competitors: both are implementations of the Single Shot MultiBox Detector (SSD) object detection algorithm, but one uses Keras and the other uses TensorFlow, meaning a user would typically choose one framework's implementation over the other.
About ssd_keras
rykov8/ssd_keras
Port of Single Shot MultiBox Detector to Keras
About SSD-Tensorflow
balancap/SSD-Tensorflow
Single Shot MultiBox Detector in TensorFlow
Implements VGG-300 and VGG-512 architectures with multi-scale feature maps and anchor box encoding/decoding for real-time object detection. Provides modular components for datasets (Pascal VOC, COCO conversion to TF-Records), network definitions, and augmentation routines, enabling fine-tuning from Caffe checkpoints or ImageNet pre-trained backbones with selective layer freezing.
Scores updated daily from GitHub, PyPI, and npm data. How scores work