A compact monocular depth estimation model fine-tuned on KITTI dataset using DINOv2 small backbone for efficient depth prediction.