REVIEW 17 cited by
ManiSkill2: A Unified Benchmark for Generalizable Manipulation Skills
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Generalizable manipulation skills, which can be composed to tackle long-horizon and complex daily chores, are one of the cornerstones of Embodied AI. However, existing benchmarks, mostly composed of a suite of simulatable environments, are insufficient to push cutting-edge research works because they lack object-level topological and geometric variations, are not based on fully dynamic simulation, or are short of native support for multiple types of manipulation tasks. To this end, we present ManiSkill2, the next generation of the SAPIEN ManiSkill benchmark, to address critical pain points often encountered by researchers when using benchmarks for generalizable manipulation skills. ManiSkill2 includes 20 manipulation task families with 2000+ object models and 4M+ demonstration frames, which cover stationary/mobile-base, single/dual-arm, and rigid/soft-body manipulation tasks with 2D/3D-input data simulated by fully dynamic engines. It defines a unified interface and evaluation protocol to support a wide range of algorithms (e.g., classic sense-plan-act, RL, IL), visual observations (point cloud, RGBD), and controllers (e.g., action type and parameterization). Moreover, it empowers fast visual input learning algorithms so that a CNN-based policy can collect samples at about 2000 FPS with 1 GPU and 16 processes on a regular workstation. It implements a render server infrastructure to allow sharing rendering resources across all environments, thereby significantly reducing memory usage. We open-source all codes of our benchmark (simulator, environments, and baselines) and host an online challenge open to interdisciplinary researchers.
Forward citations
Cited by 17 Pith papers
-
Ego2Robot: Scalable Robot Data Synthesis from Egocentric Human Data
Pretraining a VLA model on 18,561 hours of robot-synthesized egocentric human video mixed with robot data improves out-of-distribution manipulation success in simulation and on a real dual-arm robot.
-
Weights or Skills? A Survey of Robot-Learning Techniques: from Action-Predicting Weights to Robots that Write their Own Skills
A taxonomy of robot learning on a weights-versus-skills axis, with a five-rung self-improvement ladder whose top cell (feedback plus memory plus search) holds only a few recent systems.
-
SoftVTBench: A Safety-Aware Visuo-Tactile Benchmark for Physically Constrained Robotic Manipulation of Deformable Objects
Success-only metrics overstate deformable-manipulation performance; tactile sensing raises Safety Success (e.g. 21.4%→35.6% on Object-Soft) while Goal Success stays comparable.
-
Pipette: An Embodied Simulation Platform, Benchmark, and Data-Efficient Augmentation Framework for Wet-Lab Robotics
With 30 demos per task, simulation replay augmentation raises SmolVLA wet-lab success from ~44% to ~75% and modestly helps π0, while ACT remains competitive as a non-VLA baseline.
-
ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation
A standardized real-robot evaluation suite with controlled diversity and OOD strata diagnoses complementary failure modes of multi-task VLAs and world-action models on reasoning-oriented manipulation.
-
Beyond Imitation: Reinforcement Learning-Based Sim-Real Co-Training for VLA Models
Adding a real-world supervised loss to simulation reinforcement learning improves real-robot success and data efficiency for VLA co-training.
-
DISCOVERSE: Efficient Robot Simulation in Complex High-Fidelity Environments
A simulation framework that combines 3D Gaussian Splatting with MuJoCo reports improved zero-shot transfer of manipulation policies from simulation to real robots.
-
ArtGS:3D Gaussian Splatting for Interactive Visual-Physical Modeling and Manipulation of Articulated Objects
ArtGS combines multi-view 3D reconstruction, language-model joint initialization, and closed-loop optimization to improve articulated object manipulation.
-
WorldEval: World Model as Real-World Robot Policies Evaluator
WorldEval conditions a video generation model on a policy's internal action embeddings (Policy2Vec) and shows generated-video success rates correlate with real-world robot success rates.
-
AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation
AnyBody is a benchmark suite that tests cross-embodiment manipulation generalization along interpolation, extrapolation, and composition axes, and finds zero-shot generalization to unseen robot bodies remains difficult.
-
Towards Trustworthy Embodied Intelligence: A Systems Framework and Graded Trustworthiness Levels
A four-layer systems framework and T0–T5 hierarchy for grading and maintaining bounded trustworthiness claims in embodied AI systems.
-
JoyNexus: Service-Oriented Multi-Tenant Post-Training for VLA Models
A service-oriented multi-tenant architecture with schema-compatible group batching reduces aggregate GPU time for VLA post-training by about 28% in simulation.
-
DIPOLE: Fusing Vision and Geometry for Robust Visuomotor Generalization
Fusing RGB and point-cloud inputs with training-time modality dropout plus cross-attention makes a diffusion visuomotor policy markedly more robust to visual and spatial shifts than unimodal or naively fused baselines.
-
Reward Evolution with Graph-of-Thoughts: A Bi-Level Language Model Framework for Reinforcement Learning
RE-GoT combines graph-of-thoughts planning in LLMs with VLM feedback from rollout videos to automatically write and refine RL reward functions, beating prior LLM-based reward design on RoboGen and ManiSkill2.
-
RoboPearls: Editable Video Simulation for Robot Manipulation
RoboPearls is a 3D Gaussian Splatting based framework that edits demonstration videos into varied photorealistic simulations, and training on them improves robot manipulation success rates on RLBench and COLOSSEUM.
-
Evaluating Uncertainty and Quality of Visual Language Action-enabled Robots
Across three VLA models and four simulated manipulation tasks, motion-instability and goal-distance metrics correlate with expert-rated execution quality, showing that binary success rates hide large quality differences.
-
Data Pyramid for Embodied Manipulation
Embodied training data form a five-layer pyramid—real-robot, UMI, ego/exo, simulation, general V–L—ordered by the trade-off between scale and robot alignment, and model capabilities track how those layers are mixed.
Discussion (0). Sign in to comment.