Enhancing 3D Transformer Segmentation Model for Medical Image with Token-level Representation Learning

In the field of medical images, although various works find Swin Transformer has promising effectiveness on pixelwise dense prediction, whether pre-training these models without using extra dataset can further boost the performance for the downstream semantic segmentation remains unexplored.Applications of previous representation learning methods are hindered by the limited number of 3D volumes and high computational cost. In addition, most of pretext tasks designed specifically for Transformer are not applicable to hierarchical structure of Swin Transformer. Thus, this work proposes a token-level representation learning loss that maximizes agreement between token embeddings from different augmented views individually instead of volume-level global features. Moreover, we identify a potential representation collapse exclusively caused by this new loss. To prevent collapse, we invent a simple "rotate-and-restore" mechanism, which rotates and flips one augmented view of input volume, and later restores the order of tokens in the feature maps. We also modify the contrastive loss to address the discrimination between tokens at the same position but from different volumes. We test our pre-training scheme on two public medical segmentation datasets, and the results on the downstream segmentation task show more improvement of our methods than other state-of-the-art pre-trainig methods.

Arm-Constrained Curriculum Learning for Loco-Manipulation of the Wheel-Legged Robot

Incorporating a robotic manipulator into a wheel-legged robot enhances its agility and expands its potential for practical applications. However, the presence of potential instability and uncertainties presents additional challenges for control objectives. In this paper, we introduce an arm-constrained curriculum learning architecture to tackle the issues introduced by adding the manipulator. Firstly, we develop an arm-constrained reinforcement learning algorithm to ensure safety and stability in control performance. Additionally, to address discrepancies in reward settings between the arm and the base, we propose a reward-aware curriculum learning method. The policy is first trained in Isaac gym and transferred to the physical robot to do dynamic grasping tasks, including the door-opening task, fan-twitching task and the relay-baton-picking and following task. The results demonstrate that our proposed approach effectively controls the arm-equipped wheel-legged robot to master dynamic grasping skills, allowing it to chase and catch a moving object while in motion. Please refer to our website (https://acodedog.github.io/wheel-legged-loco-manipulation) for the code and supplemental videos.

Gem5Pred: Predictive Approaches For Gem5 Simulation Time

Gem5, an open-source, flexible, and cost-effective simulator, is widely recognized and utilized in both academic and industry fields for hardware simulation. However, the typically time-consuming nature of simulating programs on Gem5 underscores the need for a predictive model that can estimate simulation time. As of now, no such dataset or model exists. In response to this gap, this paper makes a novel contribution by introducing a unique dataset specifically created for this purpose. We also conducted analysis of the effects of different instruction types on the simulation time in Gem5. After this, we employ three distinct models leveraging CodeBERT to execute the prediction task based on the developed dataset. Our superior regression model achieves a Mean Absolute Error (MAE) of 0.546, while our top-performing classification model records an Accuracy of 0.696. Our models establish a foundation for future investigations on this topic, serving as benchmarks against which subsequent models can be compared. We hope that our contribution can simulate further research in this field. The dataset we used is available at https://github.com/XueyangLiOSU/Gem5Pred.