Hallucinating

Abstract: Given a grayscale photograph as input, this paper attacks the problem of hallucinating a plausible color version of the photograph. This problem is clearly underconstrained, so previous approaches have either relied on significant user interaction or resulted in desaturated colorizations. We propose a fully automatic approach that produces vibrant and realistic colorizations. We embrace the underlying uncertainty of the problem by posing it as a classification task and use class-rebalancing at training time to increase the diversity of colors in the result. The system is implemented as a feed-forward pass in a CNN at test time and is trained on over a million color images. We evaluate our algorithm using a "colorization Turing test," asking human participants to choose between a generated and ground truth color image. Our method successfully fools humans on 32% of the trials, significantly higher than previous methods. Moreover, we show that colorization can be a powerful pretext task for self-supervised feature learning, acting as a cross-channel encoder. This approach results in state-of-the-art performance on several feature learning benchmarks.

Abstract: Natural Language Generation (NLG) has improved exponentially in recent years thanks to the development of sequence-to-sequence deep learning technologies such as Transformer-based language models. This advancement has led to more fluent and coherent NLG, leading to improved development in downstream tasks such as abstractive summarization, dialogue generation, and data-to-text generation. However, it is also apparent that deep learning based generation is prone to hallucinate unintended text, which degrades the system performance and fails to meet user expectations in many real-world scenarios. To address this issue, many studies have been presented in measuring and mitigating hallucinated texts, but these have never been reviewed in a comprehensive manner before. In this survey, we thus provide a broad overview of the research progress and challenges in the hallucination problem in NLG. The survey is organized into two parts: (1) a general overview of metrics, mitigation methods, and future directions, and (2) an overview of task-specific research progress on hallucinations in the following downstream tasks, namely abstractive summarization, dialogue generation, generative question answering, data-to-text generation, and machine translation. This survey serves to facilitate collaborative efforts among researchers in tackling the challenge of hallucinated texts in NLG.

Abstract: Low-shot visual learning–the ability to recognize novel object categories from very few examples–is a hallmark of human visual intelligence. Existing machine learning approaches fail to generalize in the same way. To make progress on this foundational problem, we present a low-shot learning benchmark on complex images that mimics challenges faced by recognition systems in the wild. We then propose (1) representation regularization techniques, and (2) techniques to hallucinate additional training examples for data-starved classes. Together, our methods improve the effectiveness of convolutional networks in low-shot learning, improving the one-shot accuracy on novel classes by 2.3× on the challenging ImageNet dataset.

Abstract: Neuroscientist V. S. Ramachandran, M.D., Ph.D., is internationally renowned not just for his bold insights about the human brain but also for the stunning simplicity of the experiments he devises to solve neurology cases that have baffled his peers (using such tools as Q-Tips, glasses of water, and mirrors). Phantoms in the Brain is a fascinating journey into the deep architecture of the mind. In the bestselling tradition of Oliver Sacks, Dr. Ramachandran and his co-author, noted New York Times science writer Sandra Blakeslee, introduce us to a range of patients suffering from strange neurological afflictions, explain how Dr. Ramachandran's evaluations reveal what actually occurs in the brain, and explore what these findings reveal about our dreams, laughter, memory, depression, body image, and language -- in short, the very things that make us human. These mesmerizing cases illuminate elusive aspects of the brain: why we think the way we do, how we reason, how we deceive ourselves, and perhaps even why we are so clever at philosophy, music, and art. Some examples: -- A blind woman unerringly reaches out to grasp a pen -- and challenges us to find the true seat of vision. -- A woman who believes that her paralyzed arm is lifting a tray of drinks offers a look at the neurology of delusion -- and a unique opportunity to test Freud's theories of denial. -- A young man who insists that his parents are impostors reveals how the brain weaves meaning from the millions of incidents that compose a life. -- A woman who hallucinates cartoon characters suffers from a disorder that may have spawned James Thurber's famed flights of visual fancy -- and illustrates how in a sense we are all hallucinating, all the time. These are stories of inspired medical detective work that push the boundaries of medicine's last great frontier -- human mind -- and bring us face-to-face with new and provocative ideas about the "big questions" of the self.