# naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets- architecture

> Architecture diagram of the GitHub repository naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-09-26). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-
- Repository: https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-

## Overview

This repository documents controlled few-shot image-classification experiments: train a full-data SimpleCNN baseline, select small training subsets using different strategies, train on those subsets, and compare results across MNIST, CIFAR-10, seeds, and subset sizes. The notebooks are the product’s experiment stages. The README documents Phase 4’s use of Phase 1 checkpoints for embedding-based selection and Phase 5’s use of Phase 1 checkpoints for forgetting-event scoring. Source excerpts were unavailable, so the graph shows only these documented relationships.

Main entry points: [`phase1-baseline/phase1-baseline-final.ipynb`](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase1-baseline/phase1-baseline-final.ipynb) (Full-data baseline), [`phase-5-hard-examples-mining via Forgetting examples/phase-5-hard-examples-mining.ipynb`](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase-5-hard-examples-mining%20via%20Forgetting%20examples/phase-5-hard-examples-mining.ipynb) (Forgetting mining), [`phase4-k-means/phase4-k-means.ipynb`](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase4-k-means/phase4-k-means.ipynb) (K-means coreset), [`REPRODUCIBILITY.md`](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/REPRODUCIBILITY.md) (Phase comparisons)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_baseline["Baseline Training"]
  node_phase1["Full-data baseline"]
end

subgraph group_selection["Subset Selection"]
  node_phase2["Random sampling"]
  node_phase3["Class-balanced selection"]
  node_phase4["K-means coreset"]
end

subgraph group_hard["Hard Mining"]
  node_phase5["Forgetting mining"]
  node_phase5diagnostic["Mining diagnostics"]
end

subgraph group_comparison["Evaluation"]
  node_comparisons["Phase comparisons<br/>[REPRODUCIBILITY.md]"]
end

node_researcher(("Experimenter"))
node_mnist[("MNIST")]
node_cifar[("CIFAR-10")]

node_researcher -->|"runs experiment"| node_phase1
node_researcher -->|"runs experiment"| node_phase2
node_researcher -->|"runs experiment"| node_phase3
node_researcher -->|"runs experiment"| node_phase4
node_researcher -->|"runs experiment"| node_phase5
node_mnist -->|"trains on"| node_phase1
node_cifar -->|"trains on"| node_phase1
node_mnist -->|"samples from"| node_phase2
node_cifar -->|"samples from"| node_phase2
node_mnist -->|"samples from"| node_phase3
node_cifar -->|"samples from"| node_phase3
node_mnist -->|"embeds images"| node_phase4
node_cifar -->|"embeds images"| node_phase4
node_phase1 -->|"provides checkpoints"| node_phase4
node_mnist -->|"scores examples"| node_phase5
node_cifar -->|"scores examples"| node_phase5
node_phase1 -->|"provides checkpoints"| node_phase5
node_phase5 -->|"examines mining"| node_phase5diagnostic
node_phase1 -->|"baseline results"| node_comparisons
node_phase2 -->|"random results"| node_comparisons
node_phase3 -->|"balanced results"| node_comparisons
node_phase4 -->|"coreset results"| node_comparisons
node_phase5 -->|"mining results"| node_comparisons

click node_phase1 "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase1-baseline/phase1-baseline-final.ipynb"
click node_phase2 "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase2-random-subsampling/phase2_random_subsampling.ipynb"
click node_phase3 "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase3-class-balanced/phase3-class-balanced.ipynb"
click node_phase4 "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase4-k-means/phase4-k-means.ipynb"
click node_phase5 "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase-5-hard-examples-mining%20via%20Forgetting%20examples/phase-5-hard-examples-mining.ipynb"
click node_phase5diagnostic "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase-5-hard-examples-mining%20via%20Forgetting%20examples/fpv-p5-hard-mining-examples.ipynb"
click node_comparisons "https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/REPRODUCIBILITY.md"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_phase1 toneBlue
class node_phase2,node_phase3,node_phase4,node_mnist,node_cifar toneAmber
class node_phase5,node_phase5diagnostic toneMint
class node_comparisons toneRose
class node_researcher toneIndigo
```

## Components

### Baseline Training

- [Full-data baseline](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase1-baseline/phase1-baseline-final.ipynb): `phase1-baseline/phase1-baseline-final.ipynb`

### Subset Selection

- [Random sampling](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase2-random-subsampling/phase2_random_subsampling.ipynb): `phase2-random-subsampling/phase2_random_subsampling.ipynb`
- [Class-balanced selection](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase3-class-balanced/phase3-class-balanced.ipynb): `phase3-class-balanced/phase3-class-balanced.ipynb`
- [K-means coreset](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase4-k-means/phase4-k-means.ipynb): `phase4-k-means/phase4-k-means.ipynb`

### Hard Mining

- [Forgetting mining](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase-5-hard-examples-mining%20via%20Forgetting%20examples/phase-5-hard-examples-mining.ipynb): `phase-5-hard-examples-mining via Forgetting examples/phase-5-hard-examples-mining.ipynb`
- [Mining diagnostics](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/phase-5-hard-examples-mining%20via%20Forgetting%20examples/fpv-p5-hard-mining-examples.ipynb): `phase-5-hard-examples-mining via Forgetting examples/fpv-p5-hard-mining-examples.ipynb`

### Evaluation

- [Phase comparisons](https://github.com/naresh-gudipalli/few-shot-image-classification-can-intelligent-data-curation-replace-large-training-datasets-/blob/main/REPRODUCIBILITY.md): `REPRODUCIBILITY.md`

### Other components

- **Experimenter**
- **MNIST**
- **CIFAR-10**

## Connections

- Experimenter → Full-data baseline: runs experiment
- Experimenter → Random sampling: runs experiment
- Experimenter → Class-balanced selection: runs experiment
- Experimenter → K-means coreset: runs experiment
- Experimenter → Forgetting mining: runs experiment
- MNIST → Full-data baseline: trains on
- CIFAR-10 → Full-data baseline: trains on
- MNIST → Random sampling: samples from
- CIFAR-10 → Random sampling: samples from
- MNIST → Class-balanced selection: samples from
- CIFAR-10 → Class-balanced selection: samples from
- MNIST → K-means coreset: embeds images
- CIFAR-10 → K-means coreset: embeds images
- Full-data baseline → K-means coreset: provides checkpoints
- MNIST → Forgetting mining: scores examples
- CIFAR-10 → Forgetting mining: scores examples
- Full-data baseline → Forgetting mining: provides checkpoints
- Forgetting mining → Mining diagnostics: examines mining
- Full-data baseline → Phase comparisons: baseline results
- Random sampling → Phase comparisons: random results
- Class-balanced selection → Phase comparisons: balanced results
- K-means coreset → Phase comparisons: coreset results
- Forgetting mining → Phase comparisons: mining results