Recent papers
Four pins across on a wide desktop, then three, then two, then one. A pin may embed a video, an official figure, or only the paper. Learned systems stay off hardware tables. GroceryVLA is a product model with no public arXiv. The catalog lists every homepage paper.
Humanoid-GPT
Zero-shot humanoid motion tracking with a GPT-style causal Transformer trained on a 2B-frame motion corpus.
Real-robot experiments use a 29-DoF Unitree G1 — not Galbot G1 and not Galbot ET1. Nearest open paper to the AstraBrain-WBC 0.5 announcement; not a proven one-to-one checkpoint identity.
LATENT
Learning Athletic Humanoid Tennis Skills from Imperfect Human Motion Data. Whole-body returns on a racket-mounted humanoid.
Deployed body is a 29-DoF Unitree G1 — not Galbot G1 and not Galbot ET1. Quantitative table and extra project clips below.
GroceryVLA
Retail, supermarket, and pharmacy shelf pick. End-to-end embodied VLA product model on Galbot G1. Not a reviewed paper; no arXiv id is assigned.
UniPart
Zero-shot language-grounded 3D part segmentation for embodied interaction. Not the CVPR 2026 UniPart generation paper by Xufan He.
Paper → body map
A paper proves only the robot it evaluates. A homepage “product model” is not a checkpoint identity. Galbot G1 and Unitree G1 stay in separate tables so form classes are not mixed. Launch architecture names that have no public checkpoint card sit outside tables. Cells use a fixed vocabulary rather than prose: arXiv is an id or no public arXiv; eval body is the exact robot the source names, or a statement that it names none; claim grade is one of paper eval (the paper evaluates on this body), homepage product (the company lists it on this body, without a paper eval), or not claimed (neither source puts the system on this body).
Evaluated or listed on Galbot G1
| System | arXiv | Weights / code | Eval body in the source | End-effector in the source | Claim grade |
|---|---|---|---|---|---|
| GraspVLA | arXiv:2505.03233 [R17] | HF shengliangd/GraspVLA [R17] | real-robot evals; robot not named in the paper [R17] | — | homepage product — homepage and founder profile call it the generalist grasp stack on Galbot G1 [R21] [P3] |
| GroceryVLA | no public arXiv [R21] | — | Galbot G1 [R21] | — | homepage product — retail / pharmacy shelf-pick product model [P3] |
| TrackVLA / TrackVLA++ | arXiv:2505.23189; arXiv:2510.07134 [R19] [R25] | — | embodied visual tracking; Galbot G1 not named in the papers [R19] | — | homepage product — profile lists product-level tracking for unmanned retail [P3] |
| NavFoM | arXiv:2509.12129 [R15] | — | cross-embodiment: quadruped, drone, vehicle, wheeled platforms [R15] | — | not claimed — no named Galbot G1 SKU [R21] |
| LDA-1B | arXiv:2602.12215 [R9] | — | Galbot G1 [R9] | parallel-jaw gripper; Sharpa 22-DoF hand [R9] | paper eval — real-world figures in the paper, not a store SKU [R9] |
| WAM-TTT | arXiv:2607.06988 [R5] | — | frozen world-action model (LDA line); robot not named in the source [R5] | — | not claimed — no public proof of a named Galbot G1 product checkpoint [R5] |
| StereoVLA | arXiv:2512.21970 [R11] | HF shengliangd/StereoVLA [R11] | manipulation VLA; robot not named in the paper [R11] | — | not claimed — weights public; not listed as a Galbot G1 SKU [R11] |
Evaluated on Unitree G1
| System | arXiv | Weights / code | Eval body in the source | End-effector in the source | Claim grade |
|---|---|---|---|---|---|
| Humanoid-GPT | arXiv:2606.03985 [R7] | GalaxyGeneralRobotics code [R7] | 29-DoF Unitree G1 [R7] | — | paper eval — nearest open paper to AstraBrain-WBC 0.5, not a proven one-to-one product checkpoint [P4] |
| LATENT | arXiv:2603.12686 [R1] | official code on Galbot GitHub [R1] | 29-DoF Unitree G1 [R1] | — | paper eval — tennis whole-body control; the announced Galbot biped’s tennis mention is not this paper’s eval [R1] |
| LDA-1B (BrainCo variant) | arXiv:2602.12215 [R9] | — | Unitree G1 [R9] | BrainCo 10-DoF hand [R9] | paper eval — not a BrainCo Revo 2 product column [R9] |
Benchmarks & special-env suites
Named datasets and evaluation suites from the same lab. A benchmark is not a robot SKU and does not enter a hardware table. Use these when a method paper’s “special env” needs a public test set: humanoid tracking, spatial VQA, open-instruction rearrangement, dex grasp, articulated parts, tracking-in-the-wild, urban micromobility.
GAPartNet
Part-centric articulated objects. Special env: drawers, doors, lids — not rigid free objects. GAPartManip (ICRA 2025) extends the manipulation set.
Paper · Project · GAPartManip [R41]
EVT-Bench
Embodied Visual Tracking Benchmark named in TrackVLA (~1.7M samples). Special env: occlusion, night, distractors, long-horizon follow — not tabletop manipulation.
EI-30k
30k+ hours of mixed-quality human and robot data (sim+real, labeled+unlabeled) used to scale LDA-1B. Special env: ingesting noisy human video next to robot trajectories, not a clean single-task demo set.
Paper catalog
One listing of the papers on the lab homepage (2026–2023 plus selected earlier lineage). This is the catalog intended as a single GitHub Issue; Issues cannot be opened from this environment, so the list lives here. Pins above are a subset. Refresh from hughw19.github.io before treating this as complete. GroceryVLA is a product model, not a paper.
2026
- WAM-TTT — preprint · arXiv:2607.06988 · pin · paper
- HumanTracker — ECCV 2026 · arXiv:2608.13555 · pin · paper
- UniPart — ECCV 2026 · no arXiv on homepage · pin
- MM-Nav — ECCV 2026 · arXiv:2510.03142 · pin · paper
- GraspADMM — IROS 2026 · arXiv:2603.13832 · pin · paper
- LATENT — IROS 2026 · arXiv:2603.12686 · pin · paper
- HDFlow — ICML 2026 spotlight · arXiv:2605.04525 · pin · paper
- LIMMT — ICML 2026 · arXiv:2606.06953 · pin · paper
- LDA-1B — RSS 2026 · arXiv:2602.12215 · pin · paper
- DAPL — RSS 2026 Best Paper Finalist · arXiv:2603.09882 · pin · paper
- StereoVLA — RSS 2026 · arXiv:2512.21970 · pin · paper
- Humanoid-GPT — CVPR 2026 · arXiv:2606.03985 · pin · paper
- Layered 4D-Rotor Gaussian Splatting — CVPR 2026 · no arXiv on homepage · project
- CL3R — ICRA 2026 · arXiv:2507.08262 · pin · paper
- NavGSim — ICRA 2026 · arXiv:2603.15186 · paper
- Robust Differentiable Collision Detection — ICRA 2026 · arXiv:2511.06267 · paper
- UrbanVLA — ICRA 2026 · arXiv:2510.23576 · pin · paper
- TrackVLA++ — ICRA 2026 · arXiv:2510.07134 · pin · paper
- OmniSpatial — ICLR 2026 · arXiv:2506.03135 · pin · paper
- NavFoM — ICLR 2026 · arXiv:2509.12129 · pin · paper
- FoldNet — RA-L & IROS 2026 · arXiv:2505.09109 · pin · paper
- DexNDM — ICLR 2026 · arXiv:2510.08556 · pin · paper
2025
- SoFar — NeurIPS 2025 spotlight · arXiv:2502.13143 · pin · paper
- TrackVLA / EVT-Bench — CoRL 2025 · arXiv:2505.23189 · pin · paper
- FetchBot — CoRL 2025 Oral · arXiv:2502.17894 · pin · paper
- GraspVLA — CoRL 2025 · arXiv:2505.03233 · pin · paper
- DexVLG — ICCV 2025 highlight · arXiv:2507.02747 · pin · paper
- DyWA — ICCV 2025 · arXiv:2503.16806 · pin · paper
- RoboHanger — RA-L · arXiv:2412.01083 · pin · paper
- Dexonomy — RSS 2025 · arXiv:2504.18829 · pin · paper
- Uni-NaVid — RSS 2025 · arXiv:2412.06224 · pin · paper
- Code-as-Monitor — CVPR 2025 · arXiv:2412.04455 · paper
- GAPartManip — ICRA 2025 · arXiv:2411.18276 · GAPartNet pin · paper
- BODex — ICRA 2025 · arXiv:2412.16490 · pin · paper
- NaVid-4D — ICRA 2025 · pin · project
- QuadWBG — ICRA 2025 · arXiv:2411.06782 · paper
- Watch Less, Feel More — ICRA 2025 · arXiv:2502.14457 · paper
- GroceryVLA — product model · no arXiv · pin
2024
- D3RoMa — CoRL 2024 · arXiv:2409.14365 · pin · paper
- DexGraspNet 2.0 — CoRL 2024 · arXiv:2410.23004 · pin · paper
- ScissorBot — CoRL 2024 · arXiv:2409.13966 · pin · paper
- Task-Oriented Dexterous Grasp — IROS 2024 · arXiv:2309.13586 · paper
- Open6DOR — IROS 2024 Oral · pin · project
- NaVid — RSS 2024 · arXiv:2402.15852 · paper
- SAGE — RSS 2024 · arXiv:2312.01307 · paper
- MaskClustering — CVPR 2024 · arXiv:2401.07745 · paper
- STOPNet — ICRA 2024 · arXiv:2310.05717 · pin · paper
- GAMMA — ICRA 2024 · arXiv:2309.15459 · paper
- ASGrasp — ICRA 2024 · arXiv:2405.05648 · pin · paper
2023 and selected earlier lineage
- UniDexGrasp++ — ICCV 2023 Oral & Best Paper Finalist · arXiv:2304.00464 · pin · paper
- GAPartNet — CVPR 2023 Highlight · arXiv:2211.05272 · pin · paper
- 3D-Aware Object Goal Navigation — CVPR 2023 · arXiv:2212.00338 · paper
- UniDexGrasp — CVPR 2023 · arXiv:2303.00938 · paper
- PartManip — CVPR 2023 · arXiv:2303.16958 · paper
- Discrete Normalizing Flows on SO(3) — CVPR 2023 · arXiv:2304.03937 · paper
- DiGA — CVPR 2023 · arXiv:2304.02222 · paper
- DexGraspNet — ICRA 2023 · arXiv:2210.02697 · pin · paper
- GraspNeRF — ICRA 2023 · arXiv:2210.06575 · paper
- DREDS — ECCV 2022 · arXiv:2208.03792 · pin · paper
- HOI4D — CVPR 2022 · arXiv:2203.01577 · paper
- CAPTRA — ICCV 2021 Oral · arXiv:2104.03437 · paper
- NOCS — CVPR 2019 Oral · arXiv:1901.02970 · paper
- EI-30k — dataset with LDA-1B · pin
LATENT — autonomous humanoid tennis research
The deployed body is a 29-DoF Unitree G1 with the right hand replaced by a racket on a 3D-printed wrist connector. This is research on the Unitree G1 body; it is not attached to Galbot G1 or Galbot ET1.
| Metric | Published result | Evidence / condition |
|---|---|---|
| System | LATENT — Learning Athletic Humanoid TEnnis skills from imperfect human motioN daTa [R1] | Paper §1 |
| Robot body | 29-DoF Unitree G1; right hand replaced by a racket on a 3D-printed wrist connector [R1] | Paper §4.3 and §5 |
| Motion data | 5 amateur players; 5 h of unedited, unannotated primitive-skill motion; 3 × 5 m capture area [R1] | Paper §3.1 |
| Control / simulation | High-level planner and low-level controller at 50 Hz; training simulation at 2,000 Hz [R1] | Paper §3 |
| Training episode | 8 incoming balls; one launch every 2 s [R1] | Training task, paper §3.3.1 |
| Real-world success | Forehand 90.90%; backhand 77.78%; forecourt 88.89%; backcourt 81.82% [R1] | Paper Table 5; successful return lands inside the opponent's court |
| Real-world sensing | External optical motion capture estimates robot 6D pose and ball state using reflective markers [R1] | Paper §4.3; not onboard-vision-only autonomy |
Official project videos
Project and full video gallery · Paper · Code
Research asset
- LATENT racket MJCF — the racket-mounted Unitree G1 model used in the paper's MuJoCo experiments, linked from the official project. github.com/GalaxyGeneralRobotics/LATENT [R2]