This paper presents a method to recognize human actions from sequences of depth maps. Specifically, we employ an action graph to model explicitly the dynamics of the actions and a bag of 3D points to characterize a set of salient postures that correspond to the nodes in the action graph. In addition, we propose a simple, but effective projection based sampling scheme to sample the bag of 3D points from the depth maps. Experimental results have shown that over 90% recognition accuracy were achieved by sampling only about 1% 3D points from the depth maps. Compared to the 2D silhouette based recognition, the recognition errors were halved. In addition, we demonstrate the potential of the bag of points posture model to deal with occlusions through simulation.

Choanoflagellates are the closest known relatives of metazoans. To discover potential molecular mechanisms underlying the evolution of metazoan multicellularity, we sequenced and analysed the genome of the unicellular choanoflagellate Monosiga brevicollis. The genome contains approximately 9,200 intron-rich genes, including a number that encode cell adhesion and signalling protein domains that are otherwise restricted to metazoans. Here we show that the physical linkages among protein domains often differ between M. brevicollis and metazoans, suggesting that abundant domain shuffling followed the separation of the choanoflagellate and metazoan lineages. The completion of the M. brevicollis genome allows us to reconstruct with increasing resolution the genomic changes that accompanied the origin of metazoans.Choanoflagellates have long fascinated evolutionary biologists for their marked similarity to the 'feeding cells' (choanocytes) of sponges and the possibility that they might represent the closest living relatives of metazoans 1,2 . Over the past decade or so, evidence supporting this relationship has accumulated from phylogenetic analyses of nuclear and mitochondrial genes [3][4][5][6] , comparative genomics between the mitochondrial genomes of choanoflagellates, sponges and other metazoans 7,8 , and the finding that choanoflagellates express homologues of metazoan signalling and adhesion genes 9-12 . Furthermore, species-rich phylogenetic analyses demonstrate that choanoflagellates are not derived from metazoans, but instead represent a distinct lineage that evolved before the origin and diversification of metazoans (Fig. 1a, Supplementary Fig. 1 and Supplementary Note 3.1) 8,13 . By virtue of their position on the tree of life, studies of choanoflagellates provide an unparallelled window into the nature of the unicellular and colonial progenitors of metazoans 14 .Choanoflagellates are abundant and globally distributed microbial eukaryotes found in marine and freshwater environments 15,16 . Like sponge choanocytes, each cell bears an apical flagellum surrounded by a distinctive collar of actin-filled microvilli, with which choanoflagellates trap bacteria and detritus (Fig. 1b). Using this highly effective means of prey capture, choanoflagellates link bacteria to higher trophic levels and thus have critical roles in oceanic carbon cycling and in the microbial food web 17,18 .More than 125 choanoflagellate species have been identified, and all species have a unicellular life-history stage. Some can also form simple colonies of equipotent cells, although these differ substantially from the obligate associations of differentiated cells in metazoans 19 . Studies of basal metazoans indicate that the ancestral metazoan was multicellular and had differentiated cell types, an epithelium, a body plan and regulated development including gastrulation. In contrast, the last common ancestor of choanoflagellates and metazoans was unicellular or possibly capable of forming simple colonies, underscoring the abundant biologi...

Recurrent neural networks (RNNs) are known to be difficult to train due to the gradient vanishing and exploding problems and thus difficult to learn long-term patterns and construct deep networks. To address these problems, this paper proposes a new type of RNNs with the recurrent connection formulated as Hadamard product, referred to as independently recurrent neural network (IndRNN), where neurons in the same layer are independent of each other and connected across layers. The gradient vanishing and exploding problems are solved in IndRNN, and thus long-term dependencies can be learned. Moreover, an IndRNN can work with non-saturated activation functions such as ReLU (rectified linear unit) and be still trained robustly. Different deeper IndRNN architectures, including the basic stacked IndRNN, residual IndRNN and densely connected IndRNN, have been investigated, all of which can be much deeper than the existing RNNs. Furthermore, IndRNN reduces the computation at each time step and can be over 10 times faster than the commonly used Long short-term memory (LSTM). Experimental results have shown that the proposed IndRNN is able to process very long sequences and construct very deep networks. Better performances have been achieved on various tasks with IndRNNs compared with the traditional RNN and LSTM.unit (GRU) [15] have been proposed to address the gradient problems. However, the use of the hyperbolic tangent and the sigmoid functions as the activation function in these variants results in gradient decay over layers. Consequently, construction and training of a deep LSTM or GRU based RNN network is practically difficult.On the other hand, the existing RNN models share the same component σ(Wx t + Uh t−1 + b) in (1), where the recurrent connection connects all the neurons through time. This makes it hard to interpret and understand the roles of each individual neuron (e.g., what patterns each neuron responds to) without considering the others. Moreover, with the recurrent connections, matrix product is performed at each time step and the computation cannot be easily paralleled, leading to a very time-consuming process when dealing with long sequences.In this paper, we propose a new type of RNN, referred to as independently recurrent neural network (IndRNN). In the proposed IndRNN, the recurrent inputs are processed with the Hadamard product as h t = σ(Wx t +u h t−1 +b). This provides a number of advantages over the traditional RNNs including:• Able to process longer sequences: the gradient vanishing and exploding problem is solved by regulating the recurrent weights, and long-term memory can be kept in order to process long sequences. Experiments have demonstrated that an IndRNN can well process sequences over 5000 steps. • Able to construct deeper networks: multiple layers of IndRNNs can be efficiently stacked, especially with skip-connection and dense connection, to increase the depth of the network. An example of 21-layer residual IndRNN and deep densely connected In-dRNN are demonstrated in the experiments. ...

This paper presents a novel unsupervised domain adaptation method for cross-domain visual recognition. We propose a unified framework that reduces the shift between domains both statistically and geometrically, referred to as Joint Geometrical and Statistical Alignment (JGSA). Specifically, we learn two coupled projections that project the source domain and target domain data into lowdimensional subspaces where the geometrical shift and distribution shift are reduced simultaneously. The objective function can be solved efficiently in a closed form. Extensive experiments have verified that the proposed method significantly outperforms several state-of-the-art domain adaptation methods on a synthetic dataset and three different real world cross-domain visual recognition tasks.

Activating mutations in the tyrosine kinase domain of receptor tyrosine kinases (RTKs) cause cancer and skeletal disorders. Comparison of the crystal structures of unphosphorylated and phosphorylated wild-type FGFR2 kinase domains with those of seven unphosphorylated pathogenic mutants reveals an autoinhibitory "molecular brake" mediated by a triad of residues in the kinase hinge region of all FGFRs. Structural analysis shows that many other RTKs, including PDGFRs, VEGFRs, KIT, CSF1R, FLT3, TEK, and TIE, are also subject to regulation by this brake. Pathogenic mutations activate FGFRs and other RTKs by disengaging the brake either directly or indirectly.