LFG

LFG checkpoint that jointly predicts dense depth / 3D points, camera pose, per-point confidence, object segmentation (7 classes), and per-pixel motion from monocular driving video — no human labels.

LFG: Self-Supervised 4D Learning from In-the-Wild Driving Videos — trained on 58M+ unlabeled dashcam driving frames with frozen teachers generating pseudo ground truth on the fly. See the project page and the inference repo.

Model details

Architecture	LFG (autoregressive transformer with future-frame prediction)
Parameters	1.22B (fp32)
Input	monocular video, 3 history frames
Encoder	DINOv2
Heads	point/depth, camera/pose, confidence, segmentation (7 classes), motion

Files

• lfg_seg_motion_1.3b.pt — minimal inference checkpoint: model_state_dict (inference weights only), config (architecture settings), global_step. Loads with torch.load(..., weights_only=True).

Usage

Use the official inference repo:

python infer.py <video-or-image-dir> \
    --checkpoint lfg_seg_motion_1.3b.pt \
    --output-dir outputs/ \
    --save-visualizations

Outputs per window: points, local_points, camera_poses, conf, segmentation, motion.