neural-to-image-retrieval
OpeningA research repository for matching invasive mouse visual cortex recordings to candidate visual stimuli, with reproducibility notes, benchmark scripts, and baseline evaluation work expanding over time.
Open-source portal for neural data intelligence
Open neural data. Build brain-machine intelligence.
SiClink builds infrastructure for invasive brain-computer interfaces, with open tools for neural data I/O, visual decoding, benchmarks, and reproducible research workflows.
Featured release
The first public entry point in SiClink's visual decoding roadmap.
We are opening the stack progressively, starting with neural-to-image retrieval from invasive mouse visual cortex recordings.
Open source areas
Public assets organized around the neural data stack.
Inspired by open-source portals from engineering-led companies, this site will grow into a navigable catalog of repositories, benchmarks, notes, and community entry points.
Neural Data I/O
Structured ingestion, preprocessing, experiment data flow, and reusable interfaces for invasive neural recordings.
PlannedVisual Decoding
Retrieval, image matching, visual reconstruction, and evaluation workflows from cortical visual signals.
OpeningOpen Benchmarks
Standardized tasks, train/test splits, metrics, baselines, leaderboards, and comparison protocols.
PlannedRepresentation Learning
Feature learning and signal representations that support retrieval, decoding, and downstream neural data tasks.
FutureReal-Time Validation
Latency-aware systems for streaming neural data, online decoding, and closed-loop research validation.
FutureResearch Tooling
Experiment notes, tutorials, reproducibility guides, and developer-facing documentation for public releases.
OngoingRoadmap
Opening the stack step by step.
The roadmap is directional and will evolve with research progress, data availability, and community feedback.
- Visual Stimulus Retrieval from Neural Signals Match invasive mouse visual cortex recordings to candidate visual stimuli. Opening
- Spike Sorting and Online Sorting Research Preview Experimental sorting pipelines for spikes, MUA, and feature confidence estimates from broadband invasive recordings. Planned
- Mouse Visual Decoding Benchmark, MouseVDB v1 Offline benchmark with standardized splits, metrics, and evaluation scripts. Planned
- MouseVDB Leaderboard Beta Reference evaluation workflow and beta leaderboard for comparing visual decoding methods. Planned
- NHP and Real-Time Visual Decoding Benchmarks Future offline, reconstruction, and latency-aware benchmarks for mouse and non-human primate recordings. Future
Research and engineering notes
Use writing to make the work reproducible, not just visible.
As releases mature, this section can become the public home for benchmark design notes, preprocessing guides, release logs, and experiment reports.
Reproducibility Notes
Environment details, dataset preparation, reproduction attempts, and known limitations for public experiments.
Coming soonBenchmark Design
Design rationale for splits, metrics, baselines, candidate pools, and evaluation protocols.
Coming soonRelease Logs
Readable summaries of new repositories, roadmap updates, documentation improvements, and community feedback.
Coming soonCommunity
Discuss public ideas. Email private collaboration details.
Use GitHub Discussions for benchmark ideas, dataset questions, roadmap feedback, and reproducibility conversations. Please keep private contact details, confidential data, and non-public collaboration materials out of public threads.