Accurately parsing (i.e., segmenting and recognizing) muscles of freely-moving animals such as Drosophila larva in light-sheet fluorescence microscopy images is necessary to study the relationship between muscle activity and animal motions. However, this task is challenging due to the large inter-class similarity and intra-class variance of muscles, as well as the in-homogeneous intensity and blurred boundaries of neighboring muscles. Existing semantic and instance segmentation methods cannot effectively overcome these challenges, resulting in poor segmentation and unreliable classification. In this work, we propose a novel framework named MuscleParseNet that explicitly utilizes sequential and spatial contexts to address these challenges. MuscleParseNet contains a deformable muscle candidate detector (D-CMD) to detect candidate muscles, and a sequential and spatial context-based fine muscle parser (SS-FMP) to refine the candidates. D-CMD boosts Mask RCNN with deformable convolutions to capture shape variations for more accurate muscle segmentation. Moreover, SS-FMP re-classifies the detected candidates by establishing a global spatial context to explicitly reflect spatial relative location, then optimizes the classification using the sequential associations of candidates in adjacent frames, which significantly improves muscle recognition accuracy. Experiments on the synchronized muscle-motion dataset of nearly freely-moving larvae show that MuscleParseNet produces promising results, outperforming state-of-the-art semantic and instance segmentation methods.

The deformability and high degree of freedom of mollusks bring challenges in mathematical modeling and synthesis of motions. Traditional analytical and statistical models are limited by either rigid skeleton assumptions or model capacity, and have difficulty in generating realistic and multi-pattern mollusk motions. In this work, we present a large-scale dynamic pose dataset of Drosophila larvae and propose a motion synthesis model named Path2Pose to generate a pose sequence given the initial poses and the subsequent guiding path. The Path2Pose model is further used to synthesize long pose sequences of various motion patterns through a recursive generation method. Evaluation analysis results demonstrate that our novel model synthesizes highly realistic mollusk motions and achieves state-of-the-art performance. Our work proves high performance of deep neural networks for mollusk motion synthesis and the feasibility of long pose sequence synthesis based on the customized body shape and guiding path.

Most previous studies mainly have focused on the analysis of structural properties of individual neuronal networks from C. elegans. In recent years, an increasing number of synapse-level neural maps, also known as biological neural networks, have been reconstructed. However, it is not clear whether there are intrinsic similarities of structural properties of biological neural networks from different brain compartments or species. To explore this issue, we collected nine connectomes at synaptic resolution including C. elegans, and analyzed their structural properties. We found that these biological neural networks possess small-world properties and modules. Excluding the Drosophila larval visual system, these networks have rich clubs. The distributions of synaptic connection strength for these networks can be fitted by the truncated pow-law distributions. Additionally, compared with the power-law model, a log-normal distribution is a better model to fit the complementary cumulative distribution function (CCDF) of degree for these neuronal networks. Moreover, we also observed that these neural networks belong to the same superfamily based on the significance profile (SP) of small subgraphs in the network. Taken together, these findings suggest that biological neural networks share intrinsic similarities in their topological structure, revealing some principles underlying the formation of biological neural networks within and across species.

In machine learning research, many emerging applications can be (re)formulated as the composition optimization problem with nonsmooth regularization penalty. To solve this problem, traditional stochastic gradient descent (SGD) algorithm and its variants either have low convergence rate or are computationally expensive. Recently, several stochastic composition gradient algorithms have been proposed, however, these methods are still inefficient and not scalable to large-scale composition optimization problem instances. To address these challenges, we propose an asynchronous parallel algorithm, named Async-ProxSCVR, which effectively combines asynchronous parallel implementation and variance reduction method. We prove that the algorithm admits the fastest convergence rate for both strongly convex and general nonconvex cases. Furthermore, we analyze the query complexity of the proposed algorithm and prove that linear speedup is accessible when we increase the number of processors. Finally, we evaluate our algorithm Async-ProxSCVR on two representative composition optimization problems including value function evaluation in reinforcement learning and sparse mean-variance optimization problem. Experimental results show that the algorithm achieves significant speedups and is much faster than existing compared methods.

Unsupervised domain adaptation methods aim to alleviate performance degradation caused by domain-shift by learning domain-invariant representations. Existing deep domain adaptation methods focus on holistic feature alignment by matching source and target holistic feature distributions, without considering local features and their multi-mode statistics. We show that the learned local feature patterns are more generic and transferable and a further local feature distribution matching enables fine-grained feature alignment. In this paper, we present a method for learning domain-invariant local feature patterns and jointly aligning holistic and local feature statistics. Comparisons to the state-of-the-art unsupervised domain adaptation methods on two popular benchmark datasets demonstrate the superiority of our approach and its effectiveness on alleviating negative transfer.

Bioimage Informatics is a growing area that aims to extract biological knowledge from microscope images of biomedical samples automatically. Its mission is vastly challenging, however, due to the complexity of diverse imaging modalities and big scales of multi-dimensional images. One major challenge is automatic image segmentation, an essential step towards high-level modeling and analysis. While progresses in deep learning have brought the goal of automation much closer to reality, creating training data for producing powerful neural networks is often laborious. To provide a shortcut for this costly step, we propose a novel two-stage generative model for simulating voxel level training data based on a specially designed objective function of preserving foreground labels. Using segmenting neurons from LM (Light Microscopy) image stacks as a testing example, we showed that segmentation networks trained by our synthetic data were able to produce satisfactory results. Unlike other simulation methods available in the field, our method can be easily extended to many other applications because it does not involve sophisticated cell models and imaging mechanisms.

Domain adaptation is an important technique to alleviate performance degradation caused by domain shift, e.g., when training and test data come from different domains. Most existing deep adaptation methods focus on reducing domain shift by matching marginal feature distributions through deep transformations on the input features, due to the unavailability of target domain labels. We show that domain shift may still exist via label distribution shift at the classifier, thus deteriorating model performances. To alleviate this issue, we propose an approximate joint distribution matching scheme by exploiting prediction uncertainty. Specifically, we use a Bayesian neural network to quantify prediction uncertainty of a classifier. By imposing distribution matching on both features and labels (via uncertainty), label distribution mismatching in source and target data is effectively alleviated, encouraging the classifier to produce consistent predictions across domains. We also propose a few techniques to improve our method by adaptively reweighting domain adaptation loss to achieve nontrivial distribution matching and stable training. Comparisons with state of the art unsupervised domain adaptation methods on three popular benchmark datasets demonstrate the superiority of our approach, especially on the effectiveness of alleviating negative transfer.

In recent years, skeleton based action recognition is becoming an increasingly attractive alternative to existing video-based approaches, beneficial from its robust and comprehensive 3D information. In this paper, we explore an unsupervised representation learning approach for the first time to capture the long-term global motion dynamics in skeleton sequences. We design a conditional skeleton inpainting architecture for learning a fixed-dimensional representation, guided by additional adversarial training strategies. We quantitatively evaluate the effectiveness of our learning approach on three well-established action recognition datasets. Experimental results show that our learned representation is discriminative for classifying actions and can substantially reduce the sequence inpainting errors.