Ak18 centernet

[Dominique468]

Here are the 13 new models adapted for akidanet18_centernet.

All have an input_size of [None,224,224,3] except for aerial_cows, which has [None,384,384,3].
The performance given is that of the quantized model.

All files (*.h5, *.md) are automatically exported from another repository, and the README.md files are generated from a template.md file from the previous review.

[dmclellandbc]

Just the resolution for aerial cows in the summary table

[GuilbaudNicolas]

Suggested change

	| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/aerial-cows) | 224 | aerial-cows | 2.43M | 8 | 30.91% | | | 16 |
	| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/aerial-cows) | 224 | Aerial-cows | 2.43M | 8 | 30.91% | | | 16 |

You should use uppercase to be consistent.

[GuilbaudNicolas]

One space is enough after all " Dataset:".

[Dominique468]

You're right!
However, the document's presentation in Markdown format doesn't reflect this unnecessary white space.

To make things easier for you, I've removed this unnecessary whitespace character from the template.md file. This way, the next time we export these files, it will be corrected.

Thank you for your feedback.

[Dominique468]

17 models were added in two batches: - a first batch of 4 models. - a second batch of 13 models. The notation in the "Dataset" column of the general README.md file has indeed changed; however, no naming convention has been defined. To be helpful, I am willing to standardize the "Dataset" column of the general README.md file, but a naming convention would need to be defined. But is this really necessary?

________________________________ De : Nicolas Guilbaud ***@***.***> Envoyé : jeudi 6 novembre 2025 14:30 À : Brainchip-Inc/models ***@***.***> Cc : Dominique Couthier ***@***.***>; Author ***@***.***> Objet : Re: [Brainchip-Inc/models] Ak18 centernet (PR #18) CAUTION: This email originated from outside the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe. @GuilbaudNicolas commented on this pull request.

________________________________ In README.md<#18 (comment)>:

@@ -39,6 +39,19 @@ In addition, some models can be evaluated directly through [Akida Cloud](https:/

| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/trail_camera)| 224 | Trail_camera | 2.43M | 8 | 84.74% | | | 6 ☁️ | | Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/road_signs) | 224 | Road_signs | 2.43M | 8 | 65.46% | | | 6 ☁️ | | Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/furniture) | 224 | Furniture | 2.43M | 8 | 79.21% | | | 6 ☁️ | +| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/aerial-cows) | 224 | aerial-cows | 2.43M | 8 | 30.91% | | | 16 | ⬇️ Suggested change -| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/aerial-cows) | 224 | aerial-cows | 2.43M | 8 | 30.91% | | | 16 | +| Vision | Detection | [AkidaNet18/CenterNet](vision/detection/akidanet18_centernet/aerial-cows) | 224 | Aerial-cows | 2.43M | 8 | 30.91% | | | 16 | You should use uppercase to be consistent.

________________________________ In vision/detection/akidanet18_centernet/furniture/README.md<#18 (comment)>:

@@ -14,7 +14,7 @@ quantizeml: 1.0.1

## References - **AkidaNet18**: Akida model API [AkidaNet18 architecture](https://doc.brainchipinc.com/api_reference/akida_models_apis.html#akida_models.akidanet18_imagenet) -- **Dataset**: [furniture detection](https://universe.roboflow.com/projet-ai/furniture-azcxt/dataset/1) +- **Dataset**: [furniture detection](https://universe.roboflow.com/projet-ai/furniture-azcxt/dataset/1) On space is enough after all " Dataset:". — Reply to this email directly, view it on GitHub<#18 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/BPK6CYRGTJNVGS5SPI4HMVL33NEQPAVCNFSM6AAAAACLJYBZRGVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTIMRYGE4DOMBSG4>. You are receiving this because you authored the thread.Message ID: ***@***.***> This email, together with any attachments, is intended only for the use of the individuals or entities to which it is addressed, and may contain confidential, proprietary, or privileged information of BrainChip or other parties. If you are not an intended recipient, you are hereby notified any dissemination, distribution, or copying of this email, and any attachments, is strictly prohibited. If you receive this email in error, please immediately notify the sender or BrainChip, Inc., at 949.784.0040, and permanently delete the original and any copy of the email and any attachments.