Decoding visual stimuli from neural responses recorded by functional Magnetic Resonance Imaging (fMRI) presents an intriguing intersection between cognitive neuroscience and machine learning, promising advancements in understanding human visual perception. However, the task is challenging due to the noisy nature of fMRI signals and the intricate pattern of brain visual representations. To mitigate these challenges, we introduce a two-phase fMRI representation learning framework. The first phase pre-trains an fMRI feature learner with a proposed Double-contrastive Mask Auto-encoder to learn denoised representations. The second phase tunes the feature learner to attend to neural activation patterns most informative for visual reconstruction with guidance from an image auto-encoder. The optimized fMRI feature learner then conditions a latent diffusion model to reconstruct image stimuli from brain activities. Experimental results demonstrate our model's superiority in generating high-resolution and semantically accurate images, substantially exceeding previous state-of-the-art methods by 39.34% in the 50-way-top-1 semantic classification accuracy. The code implementations is available at https://github.com/soinx0629/vis

In previous columns, I've explained that there's a lot of hype surrounding the incremental improvements of the decades-old programming techniques collectively identified under the marketing buzzword "Artificial Intelligence" aka "AI." What's NOT hype is that those programming techniques (pattern recognition, neutral nets, ect.) have gotten incrementally more effective than they were in the past at playing games and performing speech recognition, automated translation, and so forth. What IS hype are the all-too-common and all-too-visible claims that AI will soon be able to perform complex tasks that involve anything resembling common sense, such as negotiating business deals, customer support and selling products. Well, maybe you'll believe a team of AI experts at Stanford University that is measuring the progress of AI. "Computers continue to lag considerably in the ability to generalize specific information into deeper meaning, [while] AI has made truly amazing strides in the past decade... computers still can't exhibit the common sense or the general intelligence of even a 5-year-old." As you're probably aware, AI is very good at playing games like poker, GO, and (most famously) chess.

This post announces Ray, a framework for efficiently running Python code on clusters and large multi-core machines. Like remote functions, actor methods return object IDs (that is, futures) that can be passed into other tasks and whose values can be retrieved with ray.get. The time required for deserialization is particularly important because one of the most common patterns in machine learning is to aggregate a large number of values (for example, neural net weights, rollouts, or other values) in a single process, so the deserialization step could happen hundreds of times in a row. To minimize the time required to deserialize objects in shared memory, we use the Apache Arrow data layout.