Tao Mine Od Images
NVIDIAOfficialCreated by NVIDIA Corporation
- Skill
- Tool integration
- Data and AI
- Apache-2.0
- Medium risk
- Has scripts
AutoGineer summary
Written by AutoGineer from the skill’s own files, not by the author
The tao-mine-od-images skill provides a structured interface for running the TAO Data Services TMM unique-neighbor matching process. It enables users to mine a bijectively-assigned set of unique source images that are most similar to a target set of samples, based on pre-computed embeddings. The skill supports both global allocation and class-stratified mining, allowing for targeted data selection in object detection workflows. It requires input parquets containing filepaths and embeddings, and produces a final list of unique filepaths along with summary statistics.
What it does
- Performs unique-neighbor matching to select source images based on embedding similarity.
- Supports global allocation for general mining tasks.
- Enables class-stratified mining to address rare classes in object detection datasets.
- Generates a final parquet file containing the selected unique source filepaths.
- Provides summary statistics including coverage percentage and per-class breakdowns.
- Offers optional visualization of per-class grids to inspect mining results.
When to use it
- When you need to select a specific number of unique source images that closely match target samples.
- When preparing training data for object detection and needing to balance class representation.
- When you have pre-computed embedding parquets and need to perform similarity-based data mining.
- When you need to exclude specific images from the source pool during the mining process.
Advantages
- Automates the selection of representative data for object detection models.
- Provides flexible allocation policies to handle both general and rare-class scenarios.
- Integrates directly with TAO Data Services for efficient GPU-accelerated processing.
- Includes validation scripts to ensure configuration integrity before execution.
Usage examples
- “Mine 500 unique images from my source pool that are most similar to my target dataset using global allocation.”
- The skill will use the provided source and target embedding parquets to identify the 500 closest unique neighbors and output them to the specified directory.
- “Perform class-stratified mining to improve representation of rare classes like person and bicycle.”
- By setting the allocation_policy to class_stratified and providing the rare_class_list and detection files, the miner will prioritize these classes during the selection process.
Install
Download the skill as a ZIP:
https://mp-staging.autogineer.com/agent-skills/nvidia/tao-mine-od-images/downloadUnzip it into the cross-client skills folder of your project:
mkdir -p .agents/skills && unzip tao-mine-od-images.zip -d .agents/skills
Documentation by NVIDIA Corporation
TAO Mine OD Images (Unique Neighbor Matching)
Use this skill to run TAO Data Services TMM unique-neighbor matching mining for object detection. The skill consumes pre-embedded source and target parquets and writes a directory of outputs including final_unique_files.parquet and summary.json. It does not compute embeddings; upstream steps must produce the source and target embedding parquets first.
The container entrypoint is:
tmm unique_neighbor_matching -e /absolute/path/to/unique_neighbor_matching.yamlInputs
The user can provide either an existing spec or the fields needed to generate one.
Required spec fields:
| Field | Meaning |
|---|---|
source_path | Absolute path to the source embeddings parquet or directory of parquets. |
target_path | Absolute path to the target embeddings parquet or directory of parquets. |
output_dir | Absolute path to the output directory. Writes final_unique_files.parquet, summary.json, and per-iteration parquets. |
desired_unique_count | Total number of unique source files to retrieve. |
Common optional fields:
| Field | Default | Meaning |
|---|---|---|
allocation_policy | global | global or class_stratified. |
distance_metric | euclidean | One of euclidean, cosine, or manhattan. Embeddings are L2-normalized before search. |
candidate_expansion_factor | 5 | Candidate-pool multiplier per iteration. Increase if desired count is not reached. |
source_embedding_column | embedding | Embedding column in source_path. |
target_embedding_column | embedding | Embedding column in target_path. |
source_filepath_column | filepath | Filepath column in source_path; also the column of final_unique_files.parquet. |
target_filepath_column | filepath | Filepath column in target_path. |
exclude_path | null | Parquet with a filepath column; those images are removed from the source pool. |
source_detection_file | null | COCO .json or KITTI label directory for the source. Required for class_stratified. |
target_detection_file | null | COCO .json or KITTI label directory for the target. Required for class_stratified. |
detection_format | null | coco or kitti. Required whenever a detection file is set; never inferred from the path. |
rare_class_list | "" | Comma-separated rare class names, e.g. "person,bicycle". Required for class_stratified. |
save_embeddings | false | Include embeddings in per-iteration parquet outputs. |
visualize | false | Save per-class visualization grids (requires Pillow and matplotlib). |
Both input parquets must contain the filepath and embedding columns. Source and target embeddings must have been produced by the same encoder; mismatched encoders produce garbage output.
The default template is assets/default_unique_neighbor_matching.yaml.
Quick Start
Run from the tao-skill-bank repo root. Resolve the pinned TAO Data Services image from versions.yaml, verify the spec, mount the run root with identical host/container paths, and stream the Docker logs.
Write the spec into the output directory. The run does not retain it, so a mined set otherwise carries no record of the budget, allocation policy or rare-class list that produced it — and those decide which images were selected. Keeping them together makes the selection recoverable from the run alone.
OUTPUT_DIR=/absolute/path/for/this/run # output_dir in the spec
SPEC="$OUTPUT_DIR/unique_neighbor_matching.yaml" # spec lives beside its outputs
RUN_ROOT=/absolute/path/that/contains/specs/data/and/results
GPU_COUNT=1
python3 skills/data/tao-mine-od-images/scripts/verify_unique_neighbor_matching_spec.py \
--spec "$SPEC"
DS_IMAGE="$(scripts/resolve_versions_key.py images.tao_toolkit.data_services)"
docker run --rm --gpus "$GPU_COUNT" --shm-size=8g --network=host \
-v "$RUN_ROOT:$RUN_ROOT" \
-w "$RUN_ROOT" \
"$DS_IMAGE" \
tmm unique_neighbor_matching -e "$SPEC"Do not pass --user $(id -u):$(id -g) to the TAO data-services container; some TAO DS images call getpass.getuser() at startup and fail when the UID is not in /etc/passwd.
Generate A Spec
If the user provides source/target paths and an output directory instead of a
ready spec, copy the template and fill in the nulls. Every tuning value it
already carries is the one this stage wants — change one only deliberately.
cp skills/data/tao-mine-od-images/assets/default_unique_neighbor_matching.yaml "$SPEC"Fill source_path, target_path, output_dir and desired_unique_count, all
as absolute paths, then validate:
python3 skills/data/tao-mine-od-images/scripts/verify_unique_neighbor_matching_spec.py --spec "$SPEC"source_path: /absolute/path/source_embeddings.parquet
target_path: /absolute/path/target_embeddings.parquet
output_dir: /absolute/path/results/mining_output
desired_unique_count: 500
allocation_policy: global # or class_stratified — see below
distance_metric: euclideanFor class-stratified mode set allocation_policy: class_stratified and supply
rare_class_list, source_detection_file, target_detection_file and
detection_format. verify rejects the policy without them: absent those
fields the miner falls back to a global match, which mines the wrong images
rather than failing.
The template is the only place a default value lives, so nothing can disagree
with it. verify reports the budget, policy and metric, since the mined parquet
is a list of filepaths and records nothing about why those files were chosen.
Keep the spec, input parquets, and output directory under RUN_ROOT so the same
paths resolve inside the container.
Preflight
Before launching Docker:
Verify Docker and GPU access:
docker info > /dev/null
nvidia-smi -LResolve and pull the data-services image if needed:
DS_IMAGE="$(scripts/resolve_versions_key.py images.tao_toolkit.data_services)"
docker image inspect "$DS_IMAGE" > /dev/null || docker pull "$DS_IMAGE"Validate the spec:
python3 skills/data/tao-mine-od-images/scripts/verify_unique_neighbor_matching_spec.py \
--spec "$SPEC"Confirm
RUN_ROOTcontains the spec, both input parquets (or directories), and the output directory. MountRUN_ROOTto the same absolute path inside Docker.
Outputs
| Artifact | Location |
|---|---|
| Mined source filepaths | output_dir/final_unique_files.parquet |
| Coverage and allocation stats | output_dir/summary.json |
| Per-iteration intermediates | output_dir/<subset>_iteration_<N>_topn_<K>.parquet |
| Per-class viz grids | output_dir/*.png (only if visualize: true) |
final_unique_files.parquet contains one filepath column. summary.json includes retrieved_unique_count, coverage_pct, and (when detection files are provided) per-class breakdowns for the target and selected source sets.
Troubleshooting
The subtask unique_neighbor_matching requires -e/--experiment_spec_file: rerun with tmm unique_neighbor_matching -e "$SPEC".
Input path not found inside Docker: use a RUN_ROOT mount where host and container paths are identical.
ValueError: detection_format is required: set detection_format: coco or detection_format: kitti whenever source_detection_file or target_detection_file is set.
ValueError: rare_class_list is required when allocation_policy is class_stratified: set rare_class_list and both detection files when using class_stratified.
Low coverage_pct in summary.json: the source pool is smaller than desired_unique_count. Expand the pool or increase candidate_expansion_factor.
No GPU or cuDF/cuML errors: mining requires at least one CUDA GPU. Check nvidia-smi -L, the Docker --gpus flag, and the NVIDIA container toolkit installation.
Files
- ATTRIBUTION.mdAdded by AutoGineer924 B
- BENCHMARK.md6.9 KiB
- LICENSELicence file added by AutoGineer11.2 KiB
- SKILL.md8.5 KiB
- assets/default_unique_neighbor_matching.yaml457 B
- config/skillspector-baseline.yaml1.3 KiB
- evals/evals.json893 B
- references/skill_info.yaml897 B
- scripts/verify_unique_neighbor_matching_spec.pyScript7.1 KiB
- skill-card.md4.8 KiB
- skill.oms.sig5.5 KiB
LICENSE
Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
Requirements and compatibility
Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta
“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.
Needs: Declares allowed tools, Needs a command-line tool, Needs Python
The author says: Requires docker, nvidia-container-toolkit, one or more CUDA GPUs, and the TAO data-services container pinned in versions.yaml.
Security
What our scanner noticed· Medium risk
- Large base64-encoded blobs
- References to credential files
- Scripts that your agent may run
This skill includes scripts that your agent may run. Read them before you use it.
A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.
Attribution and licence
- Created by
- NVIDIA Corporation
- Publisher
- NVIDIA (Official)
- Source
- NVIDIA/skills · skills/tao-mine-od-images · commit fce3cca
- Licence
- Apache License 2.0 (redistribution allowed)
- LICENSE (added by AutoGineer from LICENSE-APACHE)
- ATTRIBUTION.md (written by AutoGineer)
- Synced
AutoGineer is not affiliated with or endorsed by NVIDIA Corporation or NVIDIA.
Download ZIP
The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download tao-mine-od-images.zip