A goal-conditioned Diffusion Policy trained on sensor datasets. The model predicts joint positions (next positions along trajectory) conditioned on the current observation (joint positions, table camera image) and a goal cartesian position.

Model Architecture

Policy Type: Diffusion Policy
Framework: LeRobot
Horizon: 16 steps
Observation Steps: 1 step (single timestep)

Inputs

observation.state: Shape (batch, 1, 7) - Joint positions (7 DOF arm)
observation.goal: Shape (batch, 1, 3) - Goal cartesian position (X, Y, Z)
observation.images.table_camera: Shape (batch, 1, 3, 480, 640) - Table camera RGB images

Outputs

action: Shape (batch, 16, 7) - Joint positions (7 DOF) for 16-step horizon (next positions along trajectory)

Note: The model outputs a full 16-step horizon. Use select_action() to get the first step (batch, 7), or predict_action_chunk() to get the full horizon (batch, 16, 7).

Normalization

Input Normalization

Images (observation.images.table_camera):

Normalize from [0, 255] to [0, 1] by dividing by 255.0
Then apply mean-std normalization using dataset statistics (handled by preprocessor)

State (observation.state):

Apply min-max normalization: (state - min) / (max - min) using dataset statistics (handled by preprocessor)

Goal (observation.goal):

Apply min-max normalization: (goal - min) / (max - min) using dataset statistics (handled by preprocessor)

Output Unnormalization

Actions (action):

Apply inverse min-max normalization: action * (max - min) + min using dataset statistics (handled by postprocessor)
Note: Actions are joint positions (not velocities) - these are the next positions the robot should move to along the trajectory

Usage

python

from lerobot.policies.diffusion.modeling_diffusion import DiffusionPolicy
from lerobot.policies.factory import make_pre_post_processors

# Load model
policy = DiffusionPolicy.from_pretrained("calebescobedo/sensor-diffusion-policy-v1")

# Load preprocessor and postprocessor from the same repo
preprocessor, postprocessor = make_pre_post_processors(
    policy_cfg=policy.config,
    pretrained_path="calebescobedo/sensor-diffusion-policy-v1"
)

# Prepare inputs
batch = {
    'observation.state': state_tensor,  # (batch, 1, 7) - raw joint positions
    'observation.goal': goal_tensor,  # (batch, 1, 3) - raw goal xyz
    'observation.images.table_camera': table_img,  # (batch, 1, 3, 480, 640) - uint8 [0,255] or float [0,1]
}

# Inference
policy.eval()
with torch.no_grad():
    batch = preprocessor(batch)  # Normalizes inputs
    actions = policy.select_action(batch)  # Returns normalized actions
    actions = postprocessor(actions)  # Unnormalizes to raw joint positions

Training Details

Training: Epoch-based (ensures all trajectories seen)
Batch Size: 64
Optimizer: Adam
Mixed Precision: Enabled (AMP)
Data Loading: Optimized with persistent file handles
Datasets:
- roboset_20260112_225816.h5 (20 trajectories)
- roboset_20260113_001336.h5 (50 trajectories)

Dataset Notes

All trajectories have the same start and end positions
Single demonstration repeated 70 times
Goal: Final cartesian position [0.454, -0.133, 0.522] (constant)

License

MIT License