Deep learning has significantly advanced state-of-the-art for 3D hand pose estimation, of which accuracy can be improved with increased amounts of labelled data. However, acquiring 3D hand pose labels can be extremely difficult. In this talk, I will present our recent two works on leveraging self-supervised learning techniques for hand pose estimation from depth map. In both works, we incorporate differentiable renderer to the network and formulate training loss as model fitting error to update network parameters. In first part of the talk, I will present our earlier work which approximates hand surface with a set of spheres. We then model the pose prior as a variational lower bound with variational auto-encoder(VAE). In second part, I will present our latest work on regressing the vertex coordinates of a hand mesh model with 2D fully convolutional network(FCN) in a single forward pass. In the first stage, the network estimates a dense correspondence field for every pixel on the image grid to the mesh grid. In the second stage, we design a differentiable operator to map features learned from the previous stage and regress a 3D coordinate map on the mesh grid. Finally, we sample from the mesh grid to recover the mesh vertices, and fit it an articulated template mesh in closed form. Without any human annotation, both works can perform competitively with strongly supervised methods. The later work will also be later extended to be compatible with MANO model.
Organizers: Dimitrios Tzionas
Neurological disorders and injuries lead to a loss of sensorimotor function in the central nervous system, which controls the musculoskeletal system. Novel systems and control methods can be employed to create neuroprostheses that restore these functions to an unprecedented degree by two major advances: (1) Long standing limitations of inertial motion tracking are overcome by novel parameter estimation and sensor fusion methods. (2) A recent extension of classic learning control methods facilitates real-time pattern adaptation in artificial muscle recruitment. We review the role of these methods in the development of biomimetic neuroprostheses and discuss their potential impact in a range of further application systems including autonomous vehicles, robotics, and multi-agent networks.
Organizers: Sebastian Trimpe
First, a short analysis of the key components of my participation in SemEval 2018, an emotion analysis contest from tweets. Namely, a transfer learning approach used for emotion classification and a context-aware attention mechanism. In my second paper, I explore how brain information can improve word representations. Neural activation models that have been proposed in the literature use a set of example words for which fMRI measurements are available in order to find a mapping between word semantics and localized neural activations. I use such models to predict neural activations on a full word lexicon. Then, I propose a cognitive computational model that estimates semantic similarity in the neural activation space and investigates the relative performance of this model for various natural language processing tasks. Finally, in my most recent work I explore cross-topic word representations. In traditional Distributional Semantic Models -like word2vec- the multiple senses of a polysemous word are conflated into a single vector space representation. In my work, I propose a DSM that learns multiple distributional representations of a word based on different topics. Moreover, we project the different topic representations in a common space and apply a smoothing technique to group redundant topic vectors.
Organizers: Soubhik Sanyal
Since Hubel and Wiesel's seminal findings in the primary visual cortex (V1) more than 50 years ago, progress in vision science has been very limited along previous frameworks and schools of thoughts on understanding vision. Have we been asking the right questions? I will show observations motivating the new path. First, a drastic information bottleneck forces the brain to process only a tiny fraction of the massive visual input information; this selection is called the attentional selection, how to select this tiny fraction is critical. Second, a large body of evidence has been accumulating to suggest that the primary visual cortex (V1) is where this selection starts, suggesting that the visual cortical areas along the visual pathway beyond V1 must be investigated in light of this selection in V1. Placing attentional selection as the center stage, a new path to understanding vision is proposed (articulated in my book "Understanding vision: theory, models, and data", Oxford University Press 2014). I will show a first example of using this new path, which aims to ask new questions and make fresh progresses. I will relate our insights to artificial vision systems to discuss issues like top-down feedbacks in hierachical processing, analysis-by-synthesis, and image understanding.
Functional polymers can be easily tailored for their interaction with living organismes. In our Group, we have worked during the last 15 years in the development of this kind of polymeric materials with different funcionalities, high biocompatibility and in different forms. In this talk, we will describe the synthesis of thermosensitive thin films that can be used to prevent biofilm formation in medical devices, the preparation of biodegradable polymers specially designed for vectors for gene transfection and a new familliy of zwitterionic polymers that are able to cross intestine mucouse for oral delivery applications. The relationship between structure-functionality- applications will be discussed for every example.
Organizers: Metin Sitti
Haptic technologies in both kinesthetic and tactile aspects benefit a brand-new opportunity to recent human-machine interactive applications. In this talk, I, who believe in that one of the essential role of a researcher is pioneering new insights and knowledge, will present my previous research topics about haptic technologies and human-machine interactive applications in two branches: laser-based mid-air haptics and sensorimotor skill learning. For the former branch, I will introduce our approach named indirect laser radiation and its application. Indirect laser radiation utilizes a laser and a light-absorbing elastic medium to evoke a tapping-like tactile sensation. For the latter, I will introduce our data-driven approach for both modeling and learning of sensorimotor skills (especially, driving) with kinesthetic assistance and artificial neural networks; I call it human-like haptic assistance. To unify two different branches of my earlier studies for exploring the feasibility of the sensory channel named "touch", I will present a general research paradigm for human-machine interactive applications to which current haptic technologies can aim in future.
Organizers: Katherine J. Kuchenbecker
Needle insertion is the most essential skill in medical care; training has to be imparted not only for physicians but also for nurses and paramedics. In most needle insertion procedures, haptic feedback from the needle is the main stimulus that novices are to be trained in. For better patient safety, the classical methods of training the haptic skills have to be replaced with simulators based on new robotic and graphics technologies. The main objective of this work is to develop analytical models of needle insertion (a special case of epidural anesthesia) including the biomechanical and psychophysical concepts that simulate the needle-tissue interaction forces in linear heterogeneous tissues and to validate the model with a series of experiments. The biomechanical and perception models were validated with experiments in two stages: with and without the human intervention. The second stage is the validation using the Turing test with two different experiments: 1) to observe the perceptual difference between the simulated and the physical phantom model, and 2) to verify the effectiveness of perceptual filter between the unfiltered and filtered model response. The results showed that the model could replicate the physical phantom tissues with good accuracy. This can be further extended to a non-linear heterogeneous model. The proposed needle/tissue interaction force models can be used more often in improving realism, performance and enabling future applications in needle simulators in heterogeneous tissue. Needle insertion training simulator was developed with the simulated models using Omni Phantom and clinical trials are conducted for the face validity and construct validity. The face validity results showed that the degree of realism of virtual environments and instruments had the overall lowest mean score and ease of usage and training in hand – eye coordination had the highest mean score. The construct validity results showed that the simulator was able to successfully differentiate force and psychomotor signatures of anesthesiologists with experiences less than 5 years and more than 5 years. For the performance index of the trainees, a novel measure, Just Controllable Difference (JCD) was proposed and a preliminary study on JCD measure is explored using two experiments for the novice. A preliminary study on the use of clinical training simulations, especially needle insertion procedure in virtual environments is emphasized on two objectives: Firstly, measures of force JND with the three fingers and secondly, comparison of these measures in Non-Immersive Virtual Reality (NIVR) to that of the Immersive Virtual Reality (IVR) using psychophysical study with the Force Matching task, Constant Stimuli method, and Isometric Force Probing stimuli. The results showed a better force JND in the IVR compared to that of the NIVR. Also, a simple state observer model was proposed to explain the improvement of force JND in the IVR. This study would quantitatively reinforce the use of the IVR for the design of various medical simulators.
Organizers: Katherine J. Kuchenbecker
I will share my vision that microbiological systems should be as programmable, interactive, accessible, constructible, and useful as our personal electronic devices. Natural multi-cellular organisms and symbiotic systems achieve complex tasks through division of labor among cells. Such systems transcend current electronics and robotics in many ways, e.g., they synthesize chemicals, generate active physical forms, and self-replicate. Harnessing these features promises significant impact for manufacturing (bioelectronics / smart materials /swarm robotics), health (tissue engineering), chemistry (pathway modularization), ecology (bioremediation), biodesign (art), and more. My lab takes a synergistic bottom-up / top-down approach to achieve such transformative applications: (1) We utilize synthetic biology and biophysics approaches to engineer and understand multi-cell bacterial assemblies. We developed the first synthetic cell-cell adhesion toolbox  and optogenetic cell-surface adhesion toolbox (‘Biofilm Lithography’) . Integration with standard synthetic biology components (e.g., for signaling, differentiation, logic) now enables a new intelligent materials paradigm that rests on versatile, modular, and composable smart particles (i.e., cells). (2) We pioneered ‘Interactive Biotechnology’ that enables humans to directly interact with living multi-cell assemblies in real-time. I will provide the rational for this interactivity, demonstrate multiple applications using phototactic Euglena cells (e.g., tangible museum exhibits , biology cloud experimentation labs , biotic video games ), and show how this technology aided the discovery of new microswimmer phototaxis control strategies . Finally, I discuss architecture and swarm programming languages for future bio-electronic devices (i.e., ‘Biotic Processing Units’ – BPUs) [7,8]. REFs:  Glass, Cell ’18;  Jin, PNAS ’18;  Lee, CHI ACM ’15;  Hossain, Nature Biotech ‘16;  Cira, PLoS Biology ‘15;  Tsang, Nature Physics ’18;  Lam LOC ‘17;  Washington, PNAS ‘19.
With the advent of technology, tactile stimuli are adopted widely in many human-computer interactions. However, their perceptual and emotional characteristics are not much studied yet. In this talk, to help in understanding these characteristics, I will introduce my perception and emotion studies, as well as my future research plan. For perceptual characteristics, I will introduce an estimation method for perceived intensity of superimposed vibrations, verbal expressions for vibrotactile stimuli, and adjectival magnitude functions. Then, I will present a vibrotactile authoring tool that utilizes the adjectival magnitude functions as an application. For affective characteristics, I will introduce my emotion studies that investigate the effects of physical parameters of vibrotactile and thermal stimuli on the emotional responses using the valence-arousal space (V-A space). Then, as an application, I will present an emotion augmenting method that changes the emotion of visual stimuli in mobile devices using tactile stimuli.
Organizers: Katherine J. Kuchenbecker
Machine learning with artificial neural networks is revolutionizing science. The most advanced challenges require discovering answers autonomously. In the domain of reinforcement learning, control strategies are improved according to a reward function. The power of neural-network-based reinforcement learning has been highlighted by spectacular recent successes such as playing Go, but its benefits for physics are yet to be demonstrated. Here, we show how a network-based "agent" can discover complete quantum-error-correction strategies, protecting a collection of qubits against noise. These strategies require feedback adapted to measurement outcomes. Finding them from scratch without human guidance and tailored to different hardware resources is a formidable challenge due to the combinatorially large search space. To solve this challenge, we develop two ideas: two-stage learning with teacher and student networks and a reward quantifying the capability to recover the quantum information stored in a multiqubit system. Beyond its immediate impact on quantum computation, our work more generally demonstrates the promise of neural-network-based reinforcement learning in physics.
Organizers: Matthias Bauer
Multi-person articulated pose tracking is an important while challenging problem in human behavior understanding. In this talk, going along the road of top-down approaches, I will introduce a decent and efficient pose tracker based on pose flows. This approach can achieve real-time pose tracking without loss of accuracy. Besides, to better understand human activities in visual contents, clothes texture and geometric details also play indispensable roles. However, extrapolating them from a single image is much more difficult than rigid objects due to its large variations in pose, shape, and cloth. I will present a two-stage pipeline to predict human bodies and synthesize human novel views from one single-view image.
Organizers: Siyu Tang