Meta-Optimization for Higher Model Generalizability in Single-Image Depth Prediction

Model generalizability to unseen datasets, concerned with in-the-wild robustness, is less studied for indoor single-image depth prediction. We leverage gradient-based meta-learning for higher generalizability on zero-shot cross-dataset inference. Unlike the most-studied image classification in meta-learning, depth is pixel-level continuous range values, and mappings from each image to depth vary widely across environments. Thus no explicit task boundaries exist. We instead propose fine-grained task that treats each RGB-D pair as a task in our meta-optimization. We first show meta-learning on limited data induces much better prior (max +29.4\%). Using meta-learned weights as initialization for following supervised learning, without involving extra data or information, it consistently outperforms baselines without the method. Compared to most indoor-depth methods that only train/ test on a single dataset, we propose zero-shot cross-dataset protocols, closely evaluate robustness, and show consistently higher generalizability and accuracy by our meta-initialization. The work at the intersection of depth and meta-learning potentially drives both research streams to step closer to practical use.

Tournament selection in zeroth-level classifier systems based on average reward reinforcement learning

As a genetics-based machine learning technique, zeroth-level classifier system (ZCS) is based on a discounted reward reinforcement learning algorithm, bucket-brigade algorithm, which optimizes the discounted total reward received by an agent but is not suitable for all multi-step problems, especially large-size ones. There are some undiscounted reinforcement learning methods available, such as R-learning, which optimize the average reward per time step. In this paper, R-learning is used as the reinforcement learning employed by ZCS, to replace its discounted reward reinforcement learning approach, and tournament selection is used to replace roulette wheel selection in ZCS. The modification results in classifier systems that can support long action chains, and thus is able to solve large multi-step problems.

Learning classifier systems with memory condition to solve non-Markov problems

In the family of Learning Classifier Systems, the classifier system XCS has been successfully used for many applications. However, the standard XCS has no memory mechanism and can only learn optimal policy in Markov environments, where the optimal action is determined solely by the state of current sensory input. In practice, most environments are partially observable environments on agent's sensation, which are also known as non-Markov environments. Within these environments, XCS either fails, or only develops a suboptimal policy, since it has no memory. In this work, we develop a new classifier system based on XCS to tackle this problem. It adds an internal message list to XCS as the memory list to record input sensation history, and extends a small number of classifiers with memory conditions. The classifier's memory condition, as a foothold to disambiguate non-Markov states, is used to sense a specified element in the memory list. Besides, a detection method is employed to recognize non-Markov states in environments, to avoid these states controlling over classifiers' memory conditions. Furthermore, four sets of different complex maze environments have been tested by the proposed method. Experimental results show that our system is one of the best techniques to solve partially observable environments, compared with some well-known classifier systems proposed for these environments.