AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

AegisFlow: Training-free Non-myopic Path-safe Guided Flow Matching

Kunpeng Liu, Boshi Zhang, Keyou You

NeurIPS 2026 · Image generation › Fast diffusion sampling · approach: Guidance and alignment for flow matching · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes AegisFlow, a training-free, non-myopic guidance method for pretrained flow matching and diffusion models that steers generative trajectories to satisfy path-safety constraints such as identity preservation and obstacle avoidance.

Problem

Steering pretrained diffusion and flow matching generative trajectories to satisfy strict constraints (e.g., identity preservation in image editing, obstacle avoidance in robotics) is hard; existing training-free guidance methods enforce path-safety but, per the abstract, rely on (truncated) myopic

Approach

Non-myopic, path-safe guidance applied training-free to flow matching sampling so that constraints hold along the whole generative trajectory rather than only at each local step. Details beyond the abstract are not available.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace myopic min-norm controller with non-myopic barrier-cost HJB controlOC-Flow on CelebA-HQ · CelebA-HQ text-guided editingCLIP: 0.304 -> 0.309Improved
replace myopic min-norm controller with non-myopic barrier-cost HJB controlOC-Flow on CelebA-HQ · CelebA-HQ text-guided editingID: 0.920 -> 0.901Hurt
add MNC safeguard to OC-FlowOC-Flow · CelebA-HQ text-guided editingID: 0.732 -> 0.920Improved
AegisFlow vs SafeFlowMatcher (MNC) on maze planningflow matching planner · Maze2D maze_large_v1average curvature: 51.97 -> 44.72Improved
AegisFlow vs SafeFlowMatcher on maze planningflow matching planner · Maze2D maze_large_v1BS2: 0.041 -> 0.071Improved
AegisFlow vs SafeFlowMatcher on maze planningflow matching planner · Maze2D maze_large_v1D4RL normalized score: 1.632 -> 1.635No clear effect
AegisFlow vs DPCC-C on robot avoidanceflow matching policy · D3IL Avoidingtimesteps to goal: 69.0 -> 62.6Improved
AegisFlow vs DPCC-C on robot avoidanceflow matching policy · D3IL Avoidingsafe success rate: 0.98 -> 0.98No clear effect
lower control cost weight lambda from 1.0 to 0.4 (ablation)AegisFlow · CelebA-HQ text-guided editingCLIP: 0.299 -> 0.316Improved
lower control cost weight lambda from 1.0 to 0.4 (ablation)AegisFlow · CelebA-HQ text-guided editingLPIPS: 0.152 -> 0.224Hurt
lower control cost weight lambda from 1.0 to 0.4 (ablation)AegisFlow · CelebA-HQ text-guided editingID: 0.903 -> 0.903No clear effect
replace MNC with HJB-based controlOC-Flow · CelebA-HQ computation timeseconds: 518.26 -> 433.81Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy