PhGS: Post-Hoc Pruning and Refinement of Single-View Feed-Forward 3D Gaussian Reconstructions
Rinto Yagawa, Han Cheng, Dieter Schmalstieg, Hideo Saito, Shohei Mori
実装難易度
Hard
推論・学習コスト
High
想定用途
生成
概要
Recent single-view feed-forward 3D Gaussian Splatting (3DGS) generation predicts a fixed number of Gaussians per camera ray, introducing severe spatial redundancy. Most existing compaction strategies target multi-view setups to exploit cross-view consistency and are incompatible with single-image models. Instead of retraining the base feed-forward network to directly output compact…
何が新しいか
Recent single-view feed-forward 3D Gaussian Splatting (3DGS) generation predicts a fixed number of Gaussians per camera ray, introducing severe spatial redundancy. Most existing compaction strategies…
何に使えるか
生成
実装情報
- Paper URL
- あり
実装チェックリスト
実装または配布ページ
要確認Paper onlyの可能性があるため再実装前提で確認してください。
一次情報リンク
OKPaper
検証しやすさ
要確認公式実装が見つからないため、論文から再実装する前提です。
計算資源
要確認学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
ライセンス
未取得配布元のLICENSE、モデルカード、Paperの利用条件を確認してください。
商用利用
未取得研究利用限定、データセット由来制限、API規約の有無を確認してください。
自社データで試すなら
製造業・材料開発のExcel/CSVデータに落とし込むための最初の手順です。
- 1まず自社データを、入力条件、目的変数、評価したい指標に分けて整理します。
- 2LightGBMやRandom Forestなどのベースラインを先に作り、この手法と比較します。
- 3評価指標はR2/RMSE、AUC、異常検知の再現率、実験回数削減率など、現場の意思決定に近いものを選びます。
- 4SHAPや特徴量重要度で、効いている因子が物理・化学・工程知識と矛盾しないか確認します。
実装難易度
Hard - 公式実装が見つからないため、論文から再実装する前提です。
必要リソース
- GPU目安: High
- データセット: 論文・リポジトリ側の指定を確認してください。
- 学習要否: 再学習や評価環境の準備が必要になる可能性があります。
- 学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
実務で使う場合の注意点
- ライセンスと商用利用条件は、Paper / GitHub / Hugging Face の配布元で確認してください。
- 精度、再現性、計算コストはデータセットや評価条件に依存します。
- 個人情報や機密データを扱う場合は、入力データの保存先と外部API利用条件を確認してください。
関連記事
FIVE-VLA: Fast and EffectIVE Autonomous Driving with Recurrent Action Memory
State-of-the-art vision-language-action models (VLA) for autonomous driving face critical limitations: excessi
Sensory Precision Inference for Multimodal Arbitration under Uncertainty
Autonomous agents operating on multisensory data cannot assume that all sensory modalities remain consistently
Digital Twins for Opinion Dynamics: A Generative LLM Framework for Social Networks
The study of opinion dynamics in social networks is one of the key challenges in computational social science
The evolution of sex for artificial intelligence: a population-genetic framework for multigenerational model populations
Some aspects of AI development resemble a population process in which models are specialised, retrained on the