Papers โ€บ Auto-Encoding Scene Graphs for Image Captioning

Auto-Encoding Scene Graphs for Image Captioning

6 Dec 2018CVPR 2019 6arXiv:1812.02378archive 2025-07-28

Xu Yang, Kaihua Tang, Hanwang Zhang, Jianfei Cai

We propose Scene Graph Auto-Encoder (SGAE) that incorporates the language inductive bias into the encoder-decoder image captioning framework for more human-like captions. Intuitively, we humans use the inductive bias to compose collocations and contextual inference in discourse. For example, when we see the relation `person on bike', it is natural to replace `on' with `ride' and infer `person riding bike on a road' even the `road' is not evident. Therefore, exploiting such bias as a language prior is expected to help the conventional encoder-decoder models less likely overfit to the dataset bias and focus on reasoning. Specifically, we use the scene graph --- a directed graph (๐’ข) where an object node is connected by adjective nodes and relationship nodes --- to represent the complex structural layout of both image (โ„) and sentence (๐’ฎ). In the textual domain, we use SGAE to learn a dictionary (๐’Ÿ) that helps to reconstruct sentences in the ๐’ฎโ†’๐’ข โ†’๐’Ÿ โ†’๐’ฎ pipeline, where ๐’Ÿ encodes the desired language prior; in the vision-language domain, we use the shared ๐’Ÿ to guide the encoder-decoder in the โ„โ†’๐’ขโ†’๐’Ÿ โ†’๐’ฎ pipeline. Thanks to the scene graph representation and shared dictionary, the inductive bias is transferred across domains in principle. We validate the effectiveness of SGAE on the challenging MS-COCO image captioning benchmark, e.g., our SGAE-based single-model achieves a new state-of-the-art $127.8$ CIDEr-D on the Karpathy split, and a competitive $125.5$ CIDEr-D (c40) on the official server even compared to other ensemble models.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

yangxuntu/SGAE officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DecoderImage CaptioningInductive BiasSentence

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change ยท a person checks every report against the paper or source before anything changes; decisions are listed on /corrections