Visual-Inertial Odometry (VIO) provides a robust localization solution for Simultaneous Localization and Mapping (SLAM) systems. Self-supervised VIO, a leading approach, has the advantage of not requiring extensive ground-truth labels. Regrettably, this method still poses challenges for robotic applications, particularly unmanned aerial vehicles (UAVs), due to its computational complexity arising from inadequate model designs. To address this bottleneck, we introduce BotVIO (where “Bot” refers to “robotics”), a Transformer-based self-supervised VIO model, offering an excellent solution to alleviate computational burdens for robotics. Our lightweight backbone combines shallow CNNs with spatial-temporal-enhanced Transformers to replace conventional architectures, while the minimalist cross-fusion module uses single-layer cross-attention to enhance multimodal interaction. Extensive experiments show that, during pose estimation, BotVIO achieves a remarkable 70.37% reduction in trainable parameters and a 74.85% decrease in inference speed, reaching up to 57.80 fps on an NVIDIA Jetson NX (10W&2CORE), while improving pose accuracy and robustness. For the benefit of the community, we make public the source code.11https://github.com/wenhuiwei-ustc/BotVIO

Applying Reinforcement Learning (RL) to sequence generation models enables the direct optimization of long-term rewards (e.g., BLEU and human feedback), but typically requires large-scale sampling over a space of action sequences. This is a computational challenge as presented by the practice of sequence generation problems, such as machine translation, where we often deal with a large action space (e.g., a vocabulary) and a long action sequence (e.g., a translation). In this work, we introduce two-stage sampling and dynamic sampling approaches to improve the sampling efficiency during training sequence generation models via RL. We experiment with our approaches on the traditional sequence generation tasks, including machine translation and abstractive summarization. Furthermore, we evaluate our approaches in RL from human feedback (RLHF) through training a large language model using the reward model. Experimental results show that the efficient sampling-based RL, referred to as ESRL, can outperform all baselines in terms of both training efficiency and memory consumption. Notably, ESRL yields consistent performance gains over the strong REINFORCE, minimum risk training, and proximal policy optimization methods. The code is available at https://github.com/wangclnlp/DeepSpeed-Chat-Extension/examples/esrl.

Video paragraph captioning task aims to generate a detailed, fluent and relevant paragraph for a given video. Prior studies often focus on isolating visual objects (potential main components in a sentence) from the overall video content. They rarely explore the latent semantic relations between objects and high-level video concepts, resulting in dull or even incorrect descriptions. To create fine-grained and contextually relevant paragraph captions, we propose a novel framework that constructs a concept graph from a commonsense knowledge base and infers richer semantic meaning from the visual objects. Moreover, we employ a Vision-Guided Concept Selection Network that incorporates an under-sentence supervision mechanism to align the external knowledge with the visual information. Through extensive experiments on ActivityNet captions and YouCook2, the effectiveness of our method is demonstrated compared to state-of-the-art methods.

Video paragraph captioning task aims at generating a fine-grained, coherent and relevant paragraph for a video. Different from the images where objects are static, the temporal states of objects are changing in videos. The dynamic information could be contributed to understanding the whole video content. Existing works rarely put focus on modeling the dynamic changing state of the objects in the videos, causing the activities occurred in videos are poorly or wrongly depicted in paragraphs. To address this problem, we propose a novel Object State Tracking Network, which can capture the temporal state change of objects. However, due to the similarity of the consecutive frames in the videos, the information of the video is redundant and noisy. We further propose a semantic alignment mechanism, and enable the sentence information to refine the visual information. Extensive experiments on ActivityNet Captions demonstrate the effectiveness of our method.

This paper presents an SSD-based Block I/O Scheduler, short for SBIOS. SBIOS fully exploits the internal parallelism to improve the system performance. It dispatches the read requests to different blocks to make full use of SSD internal parallelism. For write requests, it tries to dispatch write requests to the same block to alleviate the block cross penalty and garbage collection overhead. The evaluation results show that compared with other I/O schedulers in the Linux kernel, SBIOS reduces the average response time significantly. Consequently, the performance of the SSD-based storage systems is improved.