AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

UniT3D: A Unified Transformer for 3D Dense Captioning and Visual Grounding

Zhenyu Chen, Ronghang Hu, Xinlei Chen, Matthias Nießner, Angel X. Chang

ICCV 2023 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 57 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
add synthetic 3D pre-training then joint fine-tuning (ablation)UniT3D · ScanRefer val groundingOverall Acc@0.5IoU: 36.49 -> 39.14Improved
add synthetic 3D pre-training then joint fine-tuning (ablation)UniT3D · ScanRefer val dense captioningCIDEr F1@0.5IoU: 26.68 -> 30.28Improved
add seq-to-seq objective jointly with bidirectional objective (from scratch) (ablation)UniT3D · ScanRefer val groundingOverall Acc@0.5IoU: 36.49 -> 37.45Improved
add bidirectional objective jointly with seq-to-seq (from scratch) (ablation)UniT3D · ScanRefer val dense captioningCIDEr F1@0.5IoU: 26.68 -> 27.28Improved
fine-tune jointly with both objectives instead of single objective after pre-training (ablation)UniT3D · ScanRefer val groundingOverall Acc@0.5IoU: 38.45 -> 39.14Improved
fine-tune jointly with both objectives instead of single objective after pre-training (ablation)UniT3D · ScanRefer val dense captioningCIDEr F1@0.5IoU: 28.13 -> 30.28Improved
pre-train on synthetic data cropped to up to 3 objects instead of 1 (ablation)UniT3D · ScanRefer val groundingOverall Acc@0.5IoU: 11.02 -> 13.51Improved
pre-train on synthetic data cropped to up to 5 objects instead of 3 (ablation)UniT3D · ScanRefer val dense captioningCIDEr F1@0.5IoU: 5.91 -> 2.81Hurt
pre-train on up to 5 object crops instead of 3 (ablation)UniT3D · ScanRefer val groundingOverall Acc@0.5IoU: 13.51 -> 11.87Hurt
add synthetic pre-training (ablation)UniT3D · ScanRefer val groundingUnique Acc@0.5IoU: 74.78 -> 73.14Hurt
add synthetic pre-training (ablation)UniT3D · ScanRefer val groundingMultiple Acc@0.5IoU: 27.60 -> 31.05Improved
add synthetic pre-training (ablation)UniT3D · ScanRefer valdetection mAP@0.5: 53.91 -> 54.03No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy