arXiv subjects

Zijian Gao

Publications and source records attributed to Zijian Gao.

27 records · Page 2Linked to original sources

arXiv2022

CLIP2TV: Align, Match and Distill for Video-Text Retrieval

Modern video-text retrieval frameworks basically consist of three parts: video encoder, text encoder and the similarity head. With the success on both visual and textual representation learning, transformer based encoders and fusion methods have also been adopted in the field of video-text retrieval. In this report, we present CLIP2TV, aiming at exploring where the critical elements lie in transformer based methods. To achieve this, We first revisit some recent works on multi-modal learning, then introduce some techniques into video-text retrieval, finally evaluate them through extensive experiments in different configurations. Notably, CLIP2TV achieves 52.9@R1 on MSR-VTT dataset, outperforming the previous SOTA result by 4.1%.

cs.CV↗

arXiv2022

Nuclear Norm Maximization Based Curiosity-Driven Learning

To handle the sparsity of the extrinsic rewards in reinforcement learning, researchers have proposed intrinsic reward which enables the agent to learn the skills that might come in handy for pursuing the rewards in the future, such as encouraging the agent to visit novel states. However, the intrinsic reward can be noisy due to the undesirable environment's stochasticity and directly applying the noisy value predictions to supervise the policy is detrimental to improve the learning performance and efficiency. Moreover, many previous studies employ $\ell^2$ norm or variance to measure the exploration novelty, which will amplify the noise due to the square operation. In this paper, we address aforementioned challenges by proposing a novel curiosity leveraging the nuclear norm maximization (NNM), which can quantify the novelty of exploring the environment more accurately while providing high-tolerance to the noise and outliers. We conduct extensive experiments across a variety of benchmark environments and the results suggest that NNM can provide state-of-the-art performance compared with previous curiosity methods. On 26 Atari games subset, when trained with only intrinsic reward, NNM achieves a human-normalized score of 1.09, which doubles that of competitive intrinsic rewards-based approaches. Our code will be released publicly to enhance the reproducibility.

cs.LG↗

arXiv2022

FPGA Acceleration of Image Reconstruction for Real-Time Photoacoustic Tomography

Photoacoustic (PA) imaging has been widely applied in both preclinical and clinical applications. With a significantly increasing number of data acquisition channels, fast and high-quality image reconstruction for real-time PA imaging is an open challenge in this community. In this paper, we propose a FPGA-accelerated method to achieve a much faster image reconstruction speed by 20~60 times compared with using CPU, with much-reduced system cost and power budget, from dozens of Watt (CPU) to 1~2 Watt (FPGA). Equivalently, the energy efficiency ratio (EER) is improved by ~1000 times. This FPGA acceleration method can be easily adapted to the most widely used algorithms, such as delay-and-sum (DAS) and its variants (e.g. DMAS, DAS-CF). We have performed in-vivo human finger experiments to demonstrate the feasibility and potential of the proposed method. To our best knowledge, this is the first study of accelerating PA image reconstruction based on FPGA platform.

physics.med-ph↗

arXiv2021

Coarse to Fine: Video Retrieval before Moment Localization

The current state-of-the-art methods for video corpus moment retrieval (VCMR) often use similarity-based feature alignment approach for the sake of convenience and speed. However, late fusion methods like cosine similarity alignment are unable to make full use of the information from both query texts and videos. In this paper, we combine feature alignment with feature fusion to promote the performance on VCMR.

cs.CV↗

arXiv2021

Photoacoustic digital brain: numerical modelling and image reconstruction via deep learning

Photoacoustic tomography (PAT) is a newly developed medical imaging modality, which combines the advantages of pure optical imaging and ultrasound imaging, owning both high optical contrast and deep penetration depth. Very recently, PAT is studied in human brain imaging. Nevertheless, while ultrasound waves are passing through the human skull tissues, the strong acoustic attenuation and aberration will happen, which causes photoacoustic signals' distortion. In this work, we use 10 magnetic resonance angiography (MRA) human brain volumes, and manually segment them to obtain the 2D human brain numerical phantoms for PAT. The numerical phantoms contain six kinds of tissues which are scalp, skull, white matter, gray matter, blood vessel and cerebral cortex. For every numerical phantom, optical properties are assigned to every kind of tissues. Then, Monte-Carlo based optical simulation is deployed to obtain the photoacoustic initial pressure. Then, we made two k-wave simulation cases: one takes inhomogeneous medium and uneven sound velocity into consideration, and the other not. Then we use the sensor data of the former one as the input of U-net, and the sensor data of the latter one as the output of U-net to train the network. We randomly choose 7 human brain PA sinograms as the training dataset and 3 human brain PA sinograms as the testing set. The testing result shows that our method could correct the skull acoustic aberration and obtain the blood vessel distribution inside the human brain satisfactorily.

physics.med-ph↗

arXiv2021

KnowSR: Knowledge Sharing among Homogeneous Agents in Multi-agent Reinforcement Learning

Recently, deep reinforcement learning (RL) algorithms have made great progress in multi-agent domain. However, due to characteristics of RL, training for complex tasks would be resource-intensive and time-consuming. To meet this challenge, mutual learning strategy between homogeneous agents is essential, which is under-explored in previous studies, because most existing methods do not consider to use the knowledge of agent models. In this paper, we present an adaptation method of the majority of multi-agent reinforcement learning (MARL) algorithms called KnowSR which takes advantage of the differences in learning between agents. We employ the idea of knowledge distillation (KD) to share knowledge among agents to shorten the training phase. To empirically demonstrate the robustness and effectiveness of KnowSR, we performed extensive experiments on state-of-the-art MARL algorithms in collaborative and competitive scenarios. The results demonstrate that KnowSR outperforms recently reported methodologies, emphasizing the importance of the proposed knowledge sharing for MARL.

cs.AI↗

arXiv2021

KnowRU: Knowledge Reusing via Knowledge Distillation in Multi-agent Reinforcement Learning

Recently, deep Reinforcement Learning (RL) algorithms have achieved dramatically progress in the multi-agent area. However, training the increasingly complex tasks would be time-consuming and resources-exhausting. To alleviate this problem, efficient leveraging the historical experience is essential, which is under-explored in previous studies as most of the exiting methods may fail to achieve this goal in a continuously variational system due to their complicated design and environmental dynamics. In this paper, we propose a method, named "KnowRU" for knowledge reusing which can be easily deployed in the majority of the multi-agent reinforcement learning algorithms without complicated hand-coded design. We employ the knowledge distillation paradigm to transfer the knowledge among agents with the goal to accelerate the training phase for new tasks, while improving the asymptotic performance of agents. To empirically demonstrate the robustness and effectiveness of KnowRU, we perform extensive experiments on state-of-the-art multi-agent reinforcement learning (MARL) algorithms on collaborative and competitive scenarios. The results show that KnowRU can outperform the recently reported methods, which emphasizes the importance of the proposed knowledge reusing for MARL.

cs.AI↗

arXiv2020

Time-reversed photoacoustic guided time-reversed ultrasonically encoded optical focusing

Deep-tissue optical imaging is a longstanding challenge limited by scattering. Both optical imaging and treatment can benefit from focusing light in deep tissue beyond one transport mean free path. Wavefront shaping based on time-reversed ultrasonically encoded (TRUE) optical focusing utilizes ultrasound focus, which is much less scattered than light in biological tissues as the 'guide star'. However, the traditional TRUE is limited by the ultrasound focusing area and pressure tagging efficiency, especially in acoustically heterogeneous medium. Even the improved version of iterative TRUE comes at a large time consumption, which limits the application of TRUE. To address this problem, we proposed a method called time-reversed photoacoustic wave guided time-reversed ultrasonically encoded (TRPA-TRUE) optical focusing by integrating ultrasonic focus guided by time-reversing PA signals, and the ultrasound modulation of diffused coherent light with optical phase conjugation (OPC), achieving dynamic focusing of light into scattering medium. Simulation results show that the focusing accuracy of the proposed method has been significantly improved compared with conventional TRUE, which is more suitable for practical applications that suffers severe acoustic distortion, e.g. transcranial optical focusing.

physics.optics↗

arXiv2019

Random forest model identifies serve strength as a key predictor of tennis match outcome

Tennis is a popular sport worldwide, boasting millions of fans and numerous national and international tournaments. Like many sports, tennis has benefitted from the popularity of rigorous record-keeping of game and player information, as well as the growth of machine learning methods for use in sports analytics. Of particular interest to bettors and betting companies alike is potential use of sports records to predict tennis match outcomes prior to match start. We compiled, cleaned, and used the largest database of tennis match information to date to predict match outcome using fairly simple machine learning methods. Using such methods allows for rapid fit and prediction times to readily incorporate new data and make real-time predictions. We were able to predict match outcomes with upwards of 80% accuracy, much greater than predictions using betting odds alone, and identify serve strength as a key predictor of match outcome. By combining prediction accuracies from three models, we were able to nearly recreate a probability distribution based on average betting odds from betting companies, which indicates that betting companies are using similar information to assign odds to matches. These results demonstrate the capability of relatively simple machine learning models to quite accurately predict tennis match outcomes.

cs.LG↗