TS-Fingerprint

How it works

Benefits

Comparisons

Highlighted claims

• TS-Fingerprint represents each medical time series with a fixed set of latent tokens rather than a variable-length token sequence. — Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization
• Its architecture uses learnable queries and cross-attention to map patch sequences into Fingerprint Tokens. — Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization
• The fixed-rank bottleneck prevents representation length from scaling with input length. — Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization
• TS-Fingerprint outperformed compared methods on average rank across datasets and metrics. — Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization
• The method was reported to be faster at inference than Medformer. — Learning Fingerprints for Medical Time Series with Redundancy-Constrained Information Maximization