AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Take A Step Back: Rethinking the Two Stages in Visual Reasoning

Mingyu Zhang, Jiting Cai, Mingyu Liu, Yue Xu, Cewu Lu, Yong-Lu Li

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: Chain-of-thought multimodal reasoning · 2 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
share the reasoner across tasks while keeping separate encoders (vs sharing both or only the encoder) (ablation)two-stage visual reasoning framework · RAVENaccuracy: 35.86 -> 55.72Improved
share the reasoner across tasks instead of separate ad-hoc models (ablation)ResNet-18 + MLP · RAVEN, CVR, SVRT, Bongard-HOI, Bongard-LOGOaccuracy: RAVEN 53.40 -> 55.72; Bongard-HOI 61.64 -> 57.65Mixed
deepen the symbol encoder (ResBlocks) beyond a task-specific inflection point (ablation)symbol encoder · RAVENaccuracy: 56.49 -> 39.12Hurt
deepen the symbol encoder from ResBlock 0 to ResBlock 8 (ablation)symbol encoder · SVRTaccuracy: 51.96 -> 87.33Improved
use Transformer reasoner instead of MLP (ablation)shared reasoner · VQAv2accuracy: 50.06 -> 32.64Hurt
use GCN reasoner instead of MLP (ablation)shared reasoner · CVR, SVRT, Bongard-LOGOaccuracy: CVR 72.40 -> 78.75Improved
use Neuro-Symbolic reasoner instead of MLP (ablation)shared reasoner · RAVENaccuracy: 55.72 -> 27.89Hurt
train the shared reasoner on more datasets from different domains (ablation)shared reasoner · Filtered-CoPhy Collision (3,000 samples)keypoint error: 13.09 -> 4.88Improved
use ImageNet-pretrained symbol encoder (ablation)symbol encoder · RAVEN, CVR, SVRTaccuracy: RAVEN 53.56 -> 52.21; CVR 72.40 -> 74.82No clear effect
replace separated encoders with CLIP as a general shared encoder (ablation)One-for-All · RAVENaccuracy: 55.72 -> 23.59Hurt
use MiniGPT-4 as shared reasoner instead of lite MLP (ablation)One-for-All · RAVENaccuracy: 55.7 -> 21.6Hurt
train One-for-All jointly rather than separately per task (ablation)One-for-All · RAVEN, CVR, SVRT, Bongard-HOIreasoning consistency (F1): RAVEN 76.94 -> 77.89Improved

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