Lucid Loop: A Virtual Deep Learning Biofeedback System for Lucid Dreaming Practice

This blog summarises a CHI’19 late breaking work paper, authored by Alexandra Kitson, Steve DiPaola, Bernhard E. Riecke. It will be presented at the ACM Conference on Computer-Human Interaction (CHI’19) during the poster sessions on May 7 at 10:20–11:00 and 15:20–16:00 and at the Interactivity Hot Desk on May 8 @ 15:20–16:00 and May 9 @ 10:20–11:00.

Simply put, lucid dreaming is the ultimate virtual reality. Lucid dreaming is knowing that one is dreaming while in a dream and can allow the dreamer to take control of the dream to experience virtually anything, whether that’s flying, facing a fear, or diving deep into consciousness. The practice of lucid dreaming can be quite involved, with no sure-fire way of inducing a lucid dream. Our research at the School of Interactive Arts and Technology, Simon Fraser University looks at how technology can potentially be a tool to help people learn to have more frequent and sustained lucid dreams in a way that is creative, playful, and intrinsically rewarding.

The concept of lucid dreaming has been around for a long time, found in Buddhist practices like Tibetan Dream Yoga. However, it was only until relatively recently that the scientific community was able to definitively show that, yes, people could actually lucid dream! This was in large part thanks to Dr. Stephen LaBerge’s work at Stanford, and now the Lucidity Institute, where he and his team asked lucid dream participants to slowly move their eyes left-right three times when they became lucid; this movement was captured with eye muscle sensors and lucid dream reports confirmed upon waking. Another key finding was that lucid dreaming can be learned, although some people can lucid dream spontaneously. With that, many different techniques and tools have emerged to help people learn to lucid dream (not an exhaustive list):

  • Dream Induced Lucid Dreams (DILD): practice mindfulness meditation and look for dream signs, things that seem bizarre or out-of-the-ordinary so you can ask yourself “am I dreaming?”
  • Mnemonic Induction of Lucid Dreams (MILD): practice dream recall, reality checks, affirmations, and visualizations
  • Wake-Initiated Lucid Dreams (WILD): while completely relaxed, focus on keeping conscious awareness as your body goes to sleep and visualize what you want to dream
  • Wake Back to Bed (WBTB): wake up after six hours of sleep for 20–60 minutes and then go back to sleep by visualizing and setting an intention to lucid dream
  • Brain Entrainment: sounds of certain frequencies, e.g., binaural beats and isochronic tones, guide your brain into a state of deep relaxation and focused awareness
  • Supplements: acetylcholine esterase inhibitors, like galantamine, can increase lucid dream occurrence, dream recall, as well as sensory vividness and complexity
  • Technology: sleep masks that give light flashes or audio to prompt lucidity while dreaming and headbands that detect when you’re in REM sleep to trigger binaural beats or other audio cues

With all of these techniques, there is no way of knowing if you are practicing correctly, and visualization can be difficult if you’ve never experienced a lucid dream before and because you can’t see someone else do it by example. Thus, we sought to find a technological solution to help people practice these proven techniques of lucid dreaming in a way that provides more concrete feedback. The two main approaches of lucid dreaming training we focus on are practicing focused awareness and visualizing becoming lucid.

Schematic of Lucid Loop. Image edited with Deep Dream and epainterly by Steve DiPaola.

To do this, we use neurofeedback with a consumer grade EEG, Muse 2, and an immersive virtual reality simulation whose visuals and audio become more lucid or clear as you focus your awareness. Our visuals and audio are based on phenomenological accounts of active and proficient lucid dreamers. The visuals are generated with deep dream and epainterly, an artistic machine learning system developed by Dr. Steve DiPaola’s lab, to help create a dream-like aesthetic that matches lucid dreaming reports and is intrinsically rewarding to interact with. The audio starts at a whisper and becomes louder and clearer as you focus your awareness to help gently guide you to a more relaxed and aware state. Our system, Lucid Loop, is still in development. We will be presenting Lucid Loop at CHI’19 during the poster sessions on Tuesday, May 7 @ 10:20–11:00 and 15:20–16:00 as well as at the Interactivity Hot Desk during the coffee breaks on Wednesday, May 8 @ 15:20–16:00 and Thursday, May 9 @ 10:20–11:00.

There currently does not exist, to the authors knowledge, any such technological system that tries to support lucid dreaming learning in this way. This is a largely unexplored research topic and that needs further investigation in these areas and their intersections:

  • Lucid Dreaming
  • Focused Awareness
  • Neurofeedback
  • Virtual Reality

Lucid dreaming holds a lot of promise for mastering fear, creativity, rehearsal, wish fulfillment, healing, and transformation. We need evidence-based techniques and tools of supporting lucid dream induction not only to bring lucid dreaming to anyone who wants to try it but also to bring a deeper understanding of sleep, lucid dreaming, and consciousness.

Read the full paper for details:

Alexandra Kitson, Steve DiPaola, Bernhard E. Riecke. (2019). Lucid Loop: A Virtual Deep Learning Biofeedback System for Lucid Dreaming Practice. In CHI ’19 Extended Abstracts of Human Factors in Computing Systems. Glasgow, UK. ACM Press. doi: 10.1145.3290607.3312952

You can follow this project on our website and blog.