The Earth, as proved by extensive scientific research, is at a critical juncture. With our rising Carbon Footprint and depletion of resources, the long-term future of the planet is in danger of an energy crisis and massive environmental changes. One of the primary culprits for this predicament is the burning of non-renewable sources of energy, which are drying up and results in massive carbon emissions. As a result, the world has been turning to renewable sources for a solution. Some of the most popular solutions are hydro, solar and wind power. The need for humankind to step up in order to prevent the damage to and save the environment is becoming more prominent, every second, because, it is the only solution available. Every day, billions of people place billions of footprints everywhere in the world, at their homes, transportation hubs, shopping malls: the list goes on. One of the facts that has been neglected is that these footsteps are a waste of potentially billions of mega joules of energy. This paper will describe a conceptual approach and a design on how to convert this wasted kinetic energy into electrical energy, providing a source of inexpensive electricity that can be utilized in low-power applications in our daily lives, as a solution to the rising carbon footprint of the Earth.
Virtual character animation control is a problem for which Reinforcement Learning (RL) is a viable approach. While current work have applied RL effectively to portray physics-based skills, social behaviours are challenging to design reward functions for, due to their lack of physical interaction with the world. On the other hand, data-driven implementations for these skills have been limited to supervised learning methods which require extensive training data and carry constraints on generalisability. In this paper, we propose RLAnimate, a novel data-driven deep RL approach to address this challenge, where we combine the strengths of RL together with an ability to learn from a motion dataset when creating agents. We formalise a mathematical structure for training agents by refining the conceptual roles of elements such as agents, environments, states and actions, in a way that leverages attributes of the character animation domain and model-based RL. An agent trained using our approach learns versatile animation dynamics to portray multiple behaviours, using an iterative RL training process, which becomes aware of valid behaviours via representations learnt from motion capture clips. We demonstrate, by training agents that portray realistic pointing and waving behaviours, that our approach requires a significantly lower training time, and substantially fewer sample episodes to be generated during training relative to state-of-the-art physics-based RL methods. Also, compared to existing supervised learning-based animation agents, RLAnimate needs a limited dataset of motion clips to generate representations of valid behaviours during training.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.