Reflex: Visually Grounded Reactive Humanoid Control

Humans effortlessly react to the world; they adeptly return a tennis shot, catch a falling cup, step aside from a passing cart, or pass a box tossed during truck unloading.

By Jia, Huang, Song +6Oct 8, 2026

Score█████░░░░░5.1

Key numbers

85.7% catch success from egocentric
87.9% when given privileged box
65% and 40% success at

Caveats

Preprint; not yet peer reviewed.

VerdictCompetent work. Briefs at most.

Read the original PDF Code

Abstract

Humans effortlessly react to the world; they adeptly return a tennis shot, catch a falling cup, step aside from a passing cart, or pass a box tossed during truck unloading. These everyday actions require anticipating a moving target and reflexively responding with the whole body within about a second. Such visual reactive whole-body control remains challenging for humanoids, which must (1) coordinate whole-body control, (2) react to object dynamics in real time, and (3) ground perception in egocentric vision. We study catching thrown boxes as a representative task that demands all three capabilities; existing systems cover only part of it, relying on end-effectors, motion capture, or marked objects. We present Reflex, a three-stage framework that learns these capabilities stage by stage while connecting them through a compact representation of object dynamics. First, a state-based policy learns whole-body catching with reinforcement learning. Second, the policy learns to infer the dynamics of a moving box from delayed and incomplete observations. Third, an egocentric RGB-D encoder learns to recover this dynamics representation directly from visual observations while keeping the whole-body controller fixed. This decomposition provides each capability with a direct training signal and makes visual learning substantially more scalable. In simulation, Reflex achieves 85.7% catch success from egocentric RGB-D, compared with 87.9% when given privileged box observations. Temporal visual history is critical: removing it reduces success by 35.6 and 39.8 points on centered and laterally offset throws. On a real Unitree G1, Reflex catches human-thrown boxes with 65% and 40% success at 2.5 and 5 m, reaching 76% and 59% of human catchers' success at the same distances.

Taoyang Jia, Weikai Huang, Linxin Song, Jared Darlington, Jieyu Zhao, Yue Wang, Jiafei Duan, Zhongzheng Ren, Ranjay Krishna

The editor's rubric

Heuristic review

Dimension	Level	Weight	What that level means
Leverage	███░░ 3	24%	A method or resource many groups across the field will adopt within a year.
Magnitude	██░░░ 2	18%	Solid incremental gain on a meaningful problem.
Evidence	██░░░ 2	14%	Limited: single setting, weak baselines, or an observational association presented as causal.
Novelty	██░░░ 2	16%	A new combination of known ideas.
Trajectory	██░░░ 2	18%	Some room to improve with obvious engineering.
Stakes	██░░░ 2	10%	Benefits a professional community (practitioners, clinicians, engineers).

Editor’s rationale

Heuristic triage from title and abstract text only, not a reading of the paper. Cues found: method (we propose); scale (scalable).

How the score was computed

rank-2026-10-07

Score█████░░░░░5.1

Score = 10 × (65% × adjusted merit / 10 + 25% × attention + 10% × freshness)

Merit: 3.7 / 10; Weighted rubric, evidence-gated.
Adjusted merit: 3.9 / 10; Shrunk toward the desk prior by editor confidence (34%).
Attention: 90%; Citations, upvotes, points, mentions.
Freshness: 36%; Half-life decay since publication.

Hugging Face upvotes121 (reference 25, via hf-daily, Oct 12, 2026, 13:49 UTC)
GitHub stars9 (reference 250, via hf-daily, Oct 12, 2026, 13:49 UTC)

The record

Reviewed by heuristic-v6 on Oct 12, 2026, 01:23 UTC. Paper type: method.
Categories: cs.RO
TOP, No.8 in the Artificial Intelligence edition of October 12, 2026.