trajloom-dense-trajectory-generation

name	trajloom-dense-trajectory-generation
title	TrajLoom: Spatiotemporal Consistency for Extended Trajectory Prediction
version	0.0.3
engine	skillxiv-v0.0.3-claude-opus-4.6
license	MIT
url	https://arxiv.org/abs/2603.22606
keywords	["Trajectory Prediction","Grid-Anchor Encoding","VAE","Flow Models","Spatiotemporal Consistency"]
description	Improve dense trajectory generation by replacing absolute coordinate encoding with grid-anchor offset encoding (reduces location variance 90%→10%), adding spatiotemporal consistency regularizers to VAE (30-40× improvement in 81-frame prediction), and using boundary-anchored fine-tuning for flow models. Effective for predicting extended trajectories (81 frames vs. prior 24-frame limits) in autonomous driving and video prediction where motion coherence across time steps is critical.
category	Component Innovation

What This Skill Does

Replace single-component trajectory encoding and loss functions with three complementary modifications: position-invariant grid-anchor encoding (eliminates location bias), spatiotemporal consistency regularizers for VAE (enforces motion realism), and on-policy boundary-anchored flow refinement (reduces long-horizon drift).

Component 1: Grid-Anchor Offset Encoding

Old component: Absolute pixel coordinates for trajectory points, which entangle motion representation with global position.

# Traditional: absolute coordinates bind motion to location
def encode_trajectory_absolute(trajectory):
    # trajectory shape: [T, P, 2] - T timesteps, P points, (x,y) coords
    return trajectory  # Motion patterns vary by grid cell

New component: Offset from fixed grid anchors removes location-dependent statistics.

# Grid-Anchor Offset Encoding
def encode_trajectory_grid_anchor(trajectory, grid_size=8):
    """
    Express each point as offset from its grid cell anchor.
    Removes location-driven variance that confuses motion learning.
    """
    T, P, _ = trajectory.shape
    offsets = torch.zeros_like(trajectory)

    for p in range(P):
        point_xy = trajectory[:, p]  # [T, 2]
        # Find grid cell for each point
        grid_cell = (point_xy / grid_size).long()
        grid_anchor = grid_cell * grid_size

        # Offset: point - anchor
        offsets[:, p] = point_xy - grid_anchor.float()

    return offsets

Impact: Reduces coordinate variance driven by location from ~90% to ~10%, making motion patterns consistent and learnable across image regions.

---

Component 2: Spatiotemporal Consistency Regularizers for VAE

Old component: Standard VAE with only pointwise reconstruction loss L_recon, ignoring temporal and spatial coherence.