Search for: All records

Mobility, power, and price points often dictate that robots do not have sufficient computing power on board to run contemporary robot algorithms at desired rates. Cloud computing providers such as AWS, GCP, and Azure offer immense computing power on demand, but tapping into that power from a robot is non-trivial. We present FogROS2, an open-source platform to facilitate cloud and fog robotics that is compatible with the emerging Robot Operating System 2 (ROS 2) standard. FogROS2 is completely redesigned and distinct from its predecessor FogROS1 in 9 ways, and has lower latency, overhead, and startup times; improved usability, and additional automa-tion, such as region and computer type selection. Additionally, FogROS2 was added to the official distribution of ROS 2, gaining performance, timing, and additional improvements associated with ROS 2. In examples, FogROS2 reduces SLAM latency by 50 %, reduces grasp planning time from 14 s to 1.2 s, and speeds up motion planning 28x. When compared to FogROS1, FogROS2 reduces network utilization by up to 3.8x, improves startup time by 63 %, and network round-trip latency by 97 %for images using video compression. The source code, examples, and documentation for FogROS2 are available at https://github.com/BerkeleyAutomation/FogROS2, and is available through the official ROS 2 repository at https://index.ros.org/p/fogros2/ 

Recent studies find existing self-supervised speech encoders contain primarily acoustic rather than semantic information. As a result, pipelined supervised automatic speech recognition (ASR) to large language model (LLM) systems achieve state-of-the-art results on semantic spoken language tasks by utilizing rich semantic representations from the LLM. These systems come at the cost of labeled audio transcriptions, which is expensive and time-consuming to obtain. We propose a taskagnostic unsupervised way of incorporating semantic information from LLMs into selfsupervised speech encoders without labeled audio transcriptions. By introducing semantics, we improve existing speech encoder spoken language understanding (SLU) performance by over 5% on intent classification (IC), with modest gains in named entity resolution (NER) and slot filling (SF), and spoken question answering (SQA) FF1 score by over 2%. Our approach, which uses no ASR data, achieves similar performance as methods trained on over 100 hours of labeled audio transcripts, demonstrating the feasibility of unsupervised semantic augmentations to existing speech encoders. 

Abstract Human cancers often re-express germline factors, yet their mechanistic role in oncogenesis and cancer progression remains unknown. Here we demonstrate that DEAD-box helicase 4 (DDX4), a germline factor and RNA helicase conserved in all multicellular organisms, contributes to increased cell motility and cisplatin-mediated drug resistance in small cell lung cancer (SCLC) cells. Proteomic analysis suggests that DDX4 expression upregulates proteins related to DNA repair and immune/inflammatory response. Consistent with these trends in cell lines, DDX4 depletion compromised in vivo tumor development while its overexpression enhanced tumor growth even after cisplatin treatment in nude mice. Further, the relatively higher DDX4 expression in SCLC patients correlates with decreased survival and shows increased expression of immune/inflammatory response markers. Taken together, we propose that DDX4 increases SCLC cell survival, by increasing the DNA damage and immune response pathways, especially under challenging conditions such as cisplatin treatment. 

Abstract BackgroundDoublecortin‐like kinase1 and 2 (DCLKs) are protein Ser/Thr kinases important for neuronal development. More recently, they are also reported to regulate plasticity such as cell proliferation and differentiation of stem cells and cancer cells, but the details of their functions in this biological context are still unclear. With an attempt to reveal the functions of DCLKs in plasticity regulation, we here used the sea urchin embryo that undergoes highly regulative development as an experimental model. ResultsWe found that both the transcripts and the proteins of DCLKs are uniformly present during early embryogenesis and with some enrichment in mesenchymal cells after gastrula stage. Knockdown of DCLKs induced general developmental delay and defects at day 2. Further, the damage on the embryo/larva induced ectopic expression of DCLKs in the ectoderm where the damage was most severe. Under a tumor‐prone or ‐suppressive condition, DCLKs expression was upregulated or downregulated, respectively, after damage. In both cases, the embryos showed severe developmental defects. ConclusionsTaken together, a transient upregulation of DCLKs appears to be involved in a damage response both during normal and abnormal development, and which could result in different phenotypes in a context dependent manner.