Cellular processes such as cell cycle progression, mitosis, apoptosis, and cell migration are characterized by well-defined events that are modulated as a function of time. Measuring these events in the context of time and its perturbation by small molecule compounds and RNAi can provide mechanistic information about cellular pathways being affected. We have used impedance-based time-dependent cell response profiling (TCRP) to measure and characterize cellular responses to antimitotic compounds or siRNAs. Our findings indicate that small molecule perturbation of mitosis leads to unique TCRP. We have further used this unique TCRP signature to screen 119 595 compound library and identified novel antimitotic compounds based on clustering analysis of the TCRPs. Importantly, 113 of the 117 hit compounds in the TCRP antimitotic cluster were confirmed as antimitotic based on independent assays, thus establishing the robust predictive nature of this profiling approach. In addition, potent and novel agents that induce mitotic arrest either by directly interfering with tubulin polymerization or by other mechanisms were identified. The TCRP approach allows for a practical and unbiased phenotypic profiling and screening tool for small molecule and RNAi perturbation of specific cellular pathways and time resolution of the TCRP approach can serve as a complement for other existing multidimensional profiling approaches.

In this paper, we tackle a challenging task named video-language segmentation. Given a video and a sentence in natural language, the goal is to segment the object or actor described by the sentence in video frames. To accurately denote a target object, the given sentence usually refers to multiple attributes, such as nearby objects with spatial relations, etc. In this paper, we propose a novel Polar Relative Positional Encoding (PRPE) mechanism that represents spatial relations in a ``linguistic'' way, i.e., in terms of direction and range. Sentence feature can interact with positional embeddings in a more direct way to extract the implied relative positional relations. We also propose parameterized functions for these positional embeddings to adapt real-value directions and ranges. With PRPE, we design a Polar Attention Module (PAM) as the basic module for vision-language fusion. Our method outperforms previous best method by a large margin of 11.4% absolute improvement in terms of mAP on the challenging A2D Sentences dataset. Our method also achieves competitive performances on the J-HMDB Sentences dataset.

No abstract

No abstract

No abstract