Search papers, labs, and topics across Lattice.
This study introduces the NEAR framework, which enhances brain-to-image retrieval by anchoring neural and visual representations, effectively addressing the limitations of traditional methods that rely on averaging multiple neural trials. The research reveals that while query noise contributes to retrieval inaccuracies, the placement of image representations also plays a crucial role, leading to the development of a dual approach that pulls noisy queries and image candidates towards a high-repetition anchor. The results demonstrate significant improvements in retrieval accuracy across multiple datasets, highlighting NEAR's potential for real-world applications with fewer stimulus repetitions.
Anchoring neural and visual representations can boost brain-to-image retrieval accuracy by over 9% with minimal stimulus repetitions, challenging the reliance on trial averaging.
Decoding visual information from brain signals probes neural representations and enables neuro-rehabilitation and dream decoding. Recent brain-to-image retrieval approaches have achieved promising performance, typically by averaging many (up to 80) neural trials per image, requiring repeated stimulus presentation that increases latency, cost, and user burden. When only one or a few repetitions are available, the retrieval accuracy drops sharply. This drop is commonly attributed to query noise because averaging suppresses noise and increases signal stability. However, we find a non-transitive alignment pattern: the low-repetition query signal and the image representation each align with the high-repetition center, but not directly with each other. This pattern shows that query noise is only part of the problem and that gallery placement also affects retrieval. We therefore propose a neural-anchor-based retrieval (NEAR) framework that treats the high-repetition center as an anchor and approaches it from both sides: a denoiser pulls the noisy query toward the true anchor, and a small network predicts each candidate's pseudo anchor from its image and pulls the image toward it. Across four datasets spanning EEG, MEG and fMRI, NEAR consistently improved retrieval in the few-repetition regime. On THINGS-EEG2, it improved 200-way Top-1 accuracy by 5.7 and 9.3 percentage points respectively, when averaging one and four repetitions. By anchoring neural and visual representations, NEAR reduces reliance on repeated acquisition and brings neural retrieval closer to real-world deployment.
