Unikernel is a promising runtime for serverless computing with its lightweight and isolated architecture. It offers a secure and efficient environment for applications. However, famous serverless frameworks like Knative have introduced heavyweight component sidecars to assist function instance deployment in a non-intrusive manner. But the sidecar not only hinders the throughput of unikernel function services but also consumes excessive memory resources. Moreover, the intricate network communication pathways among various services pose significant challenges for deploying unikernels in production serverless environments. Although shared-memory based communication on the same server can solve the communication bottleneck of unikernel-based function instances. The situation where malicious programs on the server make the shared memory untrustworthy limits the deployment of such technologies.We propose UKFaaS, a lightweight and high-performance serverless framework. UKFaaS leverages the advantages of customized operating systems through unikernel and it non-intrusively integrates sidecar functionality into the unikernel, avoiding the overhead of sidecar request forwarding. Additionally, UKFaaS innovatively implements data communication between unikernels in the same server to eliminate VM-Exit bottlenecks in RPC (remote process call) based on VMFUNC without relying on memory sharing. The preliminary experimental results indicate that UKFaaS can realize 1.8×-3.5× request throughput per second (RPS) compared with the advanced serverless system FaasFlow, UaaF and Nightcore in the Google online boutique microservice benchmark.

Deep learning tasks (DLT) include training and inference tasks, where training DLTs have requirements on minimizing average job completion time (JCT) and inference tasks need sufficient GPUs to meet real-time performance. Unfortunately, existing work separately deploys multi-tenant training and inference GPU cluster, leading to the high JCT of training DLTs with limited GPUs when the inference cluster is under insufficient GPU utilization due to the periodic inference workload. DeepBoot solves the challenges by utilizing idle GPUs in the inference cluster for the training DLTs. Specifically, 1) DeepBoot designsadaptive task scaling(ATS) algorithm to allocate GPUs in the training and inference clusters for training DLTs and minimize the performance loss when reclaiming inference GPUs. 2) DeepBoot implementsauto-fast elastic(AFE) training based on Pollux to reduce the restart overhead by inference GPU reclaiming. Our implementation on the testbed and large-scale simulation in Microsoft deep learning workload shows that DeepBoot can achieve 32% and 38% average JCT reduction respectively compared with the scheduler without utilizing idle GPUs in the inference cluster.

When segmenting the continuous proliferation of unlabeled remotely sensed images, unsupervised domain adaptation (UDA) has become one of the most critical techniques and achieved significant performance. But in fact, there still exists a large performance gap between the existing UDA frameworks and supervised learning methods, for the majority of UDA frameworks don’t consider the intra-domain gap in the target domain. In this paper, to further minimize the complex intra-domain shift within the target domain in remote sensing, we propose a novel iterative intra-domain adaptation framework (IterDANet), which conducts inter-domain adaptation (InterDA), entropy-based ranking (ER) and iterative intra-domain adaptation (IntraDA). Specifically, first, to enhance the performance of InterDA built upon GAN-based image-to-image translation, we propose a new generator selection strategy to assess and choose a well-trained generator for the inter-domain classifier. Then, to produce more accurate pseudo labels for IntraDA, we propose a new pseudo label generation strategy to remove both high-entropy and low-confident pixels in predicted maps of inter-domain classifier. Finally, to better reduce the intra-domain gap, we propose to cluster all the target images into multiple subdomains using ER and iteratively align the cleanest subdomain with other noisy subdomains. The extensive experiments on the benchmark dataset, which includes cross-city aerial images, highlight the superiority and effectiveness of our IterDANet against the state-of-the-art UDA frameworks.

Instance segmentation is an important method for high-resolution remote sensing images (HRRSIs) analysis. Traditional instance segmentation algorithms are not suitable to analyze complex HRRSIs that exhibit: 1) various shapes and sizes of targets; 2) a large number of small targets; and 3) data with long tail distribution. Here we introduce DB-BlendMask, an efficient and accurate instance segmentation method that can accommodate complex HRRSIs. It is composed of size balance coefficient (SBC), class balance module (CBM), and decomposed attention blender module (DA-Blender module). SBC consists of a fair weight allocation strategy for positive samples in object detection. CBM combines classification obtained in object detection stage to guide the semantic feature extraction. Complementary to a traditional convolutional neural network (CNN) architecture, DA-Blender module has the ability to considerably compress space complexity of attention and merge attention with semantic feature to generate the instance mask. We compare the performance of DB-BlendMask with a benchmark Mask R-CNN on two typical datasets, iSAID, and ISPRS Postdam. We obtain an average detection precision of 39.2% on iSAID and 63.6% on ISPRS Postdam, which corresponds to an improvement of 2.5% and 2.7%, respectively, compared to the benchmark in a real-time scenario.