Research map / Learning theory and algorithms / Prediction and decision losses
Bio-Inspired Hashing for Unsupervised Similarity Search
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| learn data-driven sparse expansive codes rather than use random FlyHash projections | BioHash · CIFAR-10, mAP@1000, k=2 | mAP@1000 (%): 14.62 -> 20.47 | Helped |
| add convolutional filters and spatial pooling | BioHash · CIFAR-10, mAP@1000, k=2 | mAP@1000 (%): 20.47 -> 26.94 | Helped |
| use sparse expansion instead of dense k-unit codes (ablation) | BioHash · CIFAR-10, mAP@1000, k=2 | mAP@1000 (%): 11.79 -> 20.47 | Helped |
| use sparse expansion instead of dense k-unit codes (ablation) | BioHash · MNIST, mAP@All, k=2 | mAP@All (%): 25.85 -> 44.38 | Helped |
| use channel inhibition with k_CI=10 rather than k_CI=1 (ablation) | BioConvHash · MNIST, mAP@All, k=2 | mAP@All (%): 56.16 -> 64.49 | Helped |
| use channel inhibition with k_CI=1 rather than k_CI=10 (ablation) | BioConvHash · CIFAR-10, mAP@1000, k=2 | mAP@1000 (%): 23.06 -> 26.94 | Helped |
| increase channel activity to k_CI=100 (ablation) | BioConvHash · MNIST, mAP@All, k=2 | mAP@All (%): 64.49 -> 23.83 | Hurt |
| add patch normalization to handle local intensity changes | BioConvHash · CIFAR-10, shadowed queries, mAP@1000 | mAP@1000 (%): 11.05 -> 26.84 | Helped |
| use learned BioHash rather than FlyHash random projections | BioHash · GloVe, cosine ground truth, mAP@100, k=16 | mAP@100 (%): 34.12 -> 76.30 | Helped |
Closest papers by meaning
- DevFly: Bio-Inspired Development of Binary Connections for Locality Preserving Sparse Codes (NeurIPS 2022 · similarity 0.90)
- Federated Nearest Neighbor Classification with a Colony of Fruit-Flies (AAAI 2022 · similarity 0.86)
- Prototypical Hash Encoding for On-the-Fly Fine-Grained Category Discovery (NeurIPS 2024 · similarity 0.85)
- Neural Embeddings for kNN Search in Biological Sequence (AAAI 2024 · similarity 0.84)
- Effective Comparative Prototype Hashing for Unsupervised Domain Adaptation (AAAI 2024 · similarity 0.83)
- KALAHash: Knowledge-Anchored Low-Resource Adaptation for Deep Hashing (AAAI 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.