Koala-36M: A Large-scale Video Dataset Improving Consistency between Fine-grained Conditions and Video Content

As visual generation technologies continue to advance, the scale of video datasets has expanded rapidly, and the quality of these datasets is critical to the performance of video generation models. We argue that temporal splitting, detailed captions, and video quality filtering are three key factors that determine dataset quality. However, existing datasets exhibit various limitations in these areas. To address these challenges, we introduce Koala-36M, a large-scale, high-quality video dataset featuring accurate temporal splitting, detailed captions, and superior video quality. The core of our approach lies in improving the consistency between fine-grained conditions and video content. Specifically, we employ a linear classifier on probability distributions to enhance the accuracy of transition detection, ensuring better temporal consistency. We then provide structured captions for the splitted videos, with an average length of 200 words, to improve text-video alignment. Additionally, we develop a Video Training Suitability Score (VTSS) that integrates multiple sub-metrics, allowing us to filter high-quality videos from the original corpus. Finally, we incorporate several metrics into the training process of the generation model, further refining the fine-grained conditions. Our experiments demonstrate the effectiveness of our data processing pipeline and the quality of the proposed Koala-36M dataset. Our dataset and code will be released at https://koala36m.github.io/.

GAP-RL: Grasps As Points for RL Towards Dynamic Object Grasping

Dynamic grasping of moving objects in complex, continuous motion scenarios remains challenging. Reinforcement Learning (RL) has been applied in various robotic manipulation tasks, benefiting from its closed-loop property. However, existing RL-based methods do not fully explore the potential for enhancing visual representations. In this letter, we propose a novel framework called Grasps As Points for RL (GAP-RL) to effectively and reliably grasp moving objects. By implementing a fast region-based grasp detector, we build a Grasp Encoder by transforming 6D grasp poses into Gaussian points and extracting grasp features as a higher-level abstraction than the original object point features. Additionally, we develop a Graspable Region Explorer for real-world deployment, which searches for consistent graspable regions, enabling smoother grasp generation and stable policy execution. To assess the performance fairly, we construct a simulated dynamic grasping benchmark involving objects with various complex motions. Experiment results demonstrate that our method effectively generalizes to novel objects and unseen dynamic motions compared to other baselines. Real-world experiments further validate the framework's sim-to-real transferability.

Absolute State-wise Constrained Policy Optimization: High-Probability State-wise Constraints Satisfaction

Enforcing state-wise safety constraints is critical for the application of reinforcement learning (RL) in real-world problems, such as autonomous driving and robot manipulation. However, existing safe RL methods only enforce state-wise constraints in expectation or enforce hard state-wise constraints with strong assumptions. The former does not exclude the probability of safety violations, while the latter is impractical. Our insight is that although it is intractable to guarantee hard state-wise constraints in a model-free setting, we can enforce state-wise safety with high probability while excluding strong assumptions. To accomplish the goal, we propose Absolute State-wise Constrained Policy Optimization (ASCPO), a novel general-purpose policy search algorithm that guarantees high-probability state-wise constraint satisfaction for stochastic systems. We demonstrate the effectiveness of our approach by training neural network policies for extensive robot locomotion tasks, where the agent must adhere to various state-wise safety constraints. Our results show that ASCPO significantly outperforms existing methods in handling state-wise constraints across challenging continuous control tasks, highlighting its potential for real-world applications.

ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AI

Simulation has enabled unprecedented compute-scalable approaches to robot learning. However, many existing simulation frameworks typically support a narrow range of scenes/tasks and lack features critical for scaling generalizable robotics and sim2real. We introduce and open source ManiSkill3, the fastest state-visual GPU parallelized robotics simulator with contact-rich physics targeting generalizable manipulation. ManiSkill3 supports GPU parallelization of many aspects including simulation+rendering, heterogeneous simulation, pointclouds/voxels visual input, and more. Simulation with rendering on ManiSkill3 can run 10-1000x faster with 2-3x less GPU memory usage than other platforms, achieving up to 30,000+ FPS in benchmarked environments due to minimal python/pytorch overhead in the system, simulation on the GPU, and the use of the SAPIEN parallel rendering system. Tasks that used to take hours to train can now take minutes. We further provide the most comprehensive range of GPU parallelized environments/tasks spanning 12 distinct domains including but not limited to mobile manipulation for tasks such as drawing, humanoids, and dextrous manipulation in realistic scenes designed by artists or real-world digital twins. In addition, millions of demonstration frames are provided from motion planning, RL, and teleoperation. ManiSkill3 also provides a comprehensive set of baselines that span popular RL and learning-from-demonstrations algorithms.

Constrained Bandwidth Observation Sharing for Multi-Robot Navigation in Dynamic Environments via Intelligent Knapsack

Multi-robot navigation is increasingly crucial in various domains, including disaster response, autonomous vehicles, and warehouse and manufacturing automation. Robot teams often must operate in highly dynamic environments and under strict bandwidth constraints imposed by communication infrastructure, rendering effective observation sharing within the system a challenging problem. This paper presents a novel optimal communication scheme, Intelligent Knapsack (iKnap), for multi-robot navigation in dynamic environments under bandwidth constraints. We model multi-robot communication as belief propagation in a graph of inferential agents. We then formulate the combinatorial optimization for observation sharing as a 0/1 knapsack problem, where each potential pairwise communication between robots is assigned a decision-making utility to be weighed against its bandwidth cost, and the system has some cumulative bandwidth limit. Compared to state-of-the-art broadcast-based optimal communication schemes, iKnap yields significant improvements in navigation performance with respect to scenario complexity while maintaining a similar runtime. Furthermore, iKnap utilizes allocated bandwidth and observational resources more efficiently than existing approaches, especially in very low-resource and high-uncertainty settings. Based on these results, we claim that the proposed method enables more robust collaboration for multi-robot teams in real-world navigation problems.

Safe Control of Quadruped in Varying Dynamics via Safety Index Adaptation

Varying dynamics pose a fundamental difficulty when deploying safe control laws in the real world. Safety Index Synthesis (SIS) deeply relies on the system dynamics and once the dynamics change, the previously synthesized safety index becomes invalid. In this work, we show the real-time efficacy of Safety Index Adaptation (SIA) in varying dynamics. SIA enables real-time adaptation to the changing dynamics so that the adapted safe control law can still guarantee 1) forward invariance within a safe region and 2) finite time convergence to that safe region. This work employs SIA on a package-carrying quadruped robot, where the payload weight changes in real-time. SIA updates the safety index when the dynamics change, e.g., a change in payload weight, so that the quadruped can avoid obstacles while achieving its performance objectives. Numerical study provides theoretical guarantees for SIA and a series of hardware experiments demonstrate the effectiveness of SIA in real-world deployment in avoiding obstacles under varying dynamics.

DexSim2Real$^{2}$: Building Explicit World Model for Precise Articulated Object Dexterous Manipulation

Articulated object manipulation is ubiquitous in daily life. In this paper, we present DexSim2Real$^{2}$, a novel robot learning framework for goal-conditioned articulated object manipulation using both two-finger grippers and multi-finger dexterous hands. The key of our framework is constructing an explicit world model of unseen articulated objects through active one-step interactions. This explicit world model enables sampling-based model predictive control to plan trajectories achieving different manipulation goals without needing human demonstrations or reinforcement learning. It first predicts an interaction motion using an affordance estimation network trained on self-supervised interaction data or videos of human manipulation from the internet. After executing this interaction on the real robot, the framework constructs a digital twin of the articulated object in simulation based on the two point clouds before and after the interaction. For dexterous multi-finger manipulation, we propose to utilize eigengrasp to reduce the high-dimensional action space, enabling more efficient trajectory searching. Extensive experiments validate the framework's effectiveness for precise articulated object manipulation in both simulation and the real world using a two-finger gripper and a 16-DoF dexterous hand. The robust generalizability of the explicit world model also enables advanced manipulation strategies, such as manipulating with different tools.

KOALA: Enhancing Speculative Decoding for LLM via Multi-Layer Draft Heads with Adversarial Learning

Large Language Models (LLMs) exhibit high inference latency due to their autoregressive decoding nature. While the draft head in speculative decoding mitigates this issue, its full potential remains unexplored. In this paper, we introduce KOALA (K-layer Optimized Adversarial Learning Architecture), an orthogonal approach to the draft head. By transforming the conventional single-layer draft head into a multi-layer architecture and incorporating adversarial learning into the traditional supervised training, KOALA significantly improves the accuracy of the draft head in predicting subsequent tokens, thus more closely mirroring the functionality of LLMs. Although this improvement comes at the cost of slightly increased drafting overhead, KOALA substantially unlocks the draft head's potential, greatly enhancing speculative decoding. We conducted comprehensive evaluations of KOALA, including both autoregressive and non-autoregressive draft heads across various tasks, demonstrating a latency speedup ratio improvement of 0.24x-0.41x, which is 10.57%-14.09% faster than the original draft heads.

ActiveNeRF: Learning Accurate 3D Geometry by Active Pattern Projection

NeRFs have achieved incredible success in novel view synthesis. However, the accuracy of the implicit geometry is unsatisfactory because the passive static environmental illumination has low spatial frequency and cannot provide enough information for accurate geometry reconstruction. In this work, we propose ActiveNeRF, a 3D geometry reconstruction framework, which improves the geometry quality of NeRF by actively projecting patterns of high spatial frequency onto the scene using a projector which has a constant relative pose to the camera. We design a learnable active pattern rendering pipeline which jointly learns the scene geometry and the active pattern. We find that, by adding the active pattern and imposing its consistency across different views, our proposed method outperforms state of the art geometry reconstruction methods qualitatively and quantitatively in both simulation and real experiments. Code is avaliable at https://github.com/hcp16/active_nerf

Topology Optimization of Random Memristors for Input-Aware Dynamic SNN

There is unprecedented development in machine learning, exemplified by recent large language models and world simulators, which are artificial neural networks running on digital computers. However, they still cannot parallel human brains in terms of energy efficiency and the streamlined adaptability to inputs of different difficulties, due to differences in signal representation, optimization, run-time reconfigurability, and hardware architecture. To address these fundamental challenges, we introduce pruning optimization for input-aware dynamic memristive spiking neural network (PRIME). Signal representation-wise, PRIME employs leaky integrate-and-fire neurons to emulate the brain's inherent spiking mechanism. Drawing inspiration from the brain's structural plasticity, PRIME optimizes the topology of a random memristive spiking neural network without expensive memristor conductance fine-tuning. For runtime reconfigurability, inspired by the brain's dynamic adjustment of computational depth, PRIME employs an input-aware dynamic early stop policy to minimize latency during inference, thereby boosting energy efficiency without compromising performance. Architecture-wise, PRIME leverages memristive in-memory computing, mirroring the brain and mitigating the von Neumann bottleneck. We validated our system using a 40 nm 256 Kb memristor-based in-memory computing macro on neuromorphic image classification and image inpainting. Our results demonstrate the classification accuracy and Inception Score are comparable to the software baseline, while achieving maximal 62.50-fold improvements in energy efficiency, and maximal 77.0% computational load savings. The system also exhibits robustness against stochastic synaptic noise of analogue memristors. Our software-hardware co-designed model paves the way to future brain-inspired neuromorphic computing with brain-like energy efficiency and adaptivity.