- Rating
- 3(+1)
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
D³-Tok predicts a content-dependent prefix length for video tokenization using complexity-based supervision and reconstruction feedback, followed by diffusion decoding.
Critical strengths
-
Addresses a practical problem: adaptive token allocation without reconstruction-based search.
-
Reports promising reconstruction and generation quality with approximately 31% fewer tokens than its fixed-length counterpart.
-
Provides component ablations and additional reconstruction metrics.
Critical weaknesses
-
Insufficient positioning against prior work. Relevant missing references include CAT[1], One-D-Piece[2], VaporTok[3], and One-DVA[4]. The contribution beyond prior complexity-aware allocation and diffusion decoding needs clarification. The characterization of KARL as requiring inference-time search is inaccurate, and comparison with EVATok is missing.
-
The benefit of content-adaptive allocation is not isolated. Comparing 1024 fixed tokens with 706 adaptive tokens cannot distinguish adaptive allocation from sequence shortening. Matched-budget fixed-length and input-independent variable-length baselines are needed.
-
The efficiency comparison is unclear. ElasticTok uses a custom sequential search with an rFVD-based stopping criterion. Since rFVD is a distribution-level metric, its application to individual videos requires explanation. Comparisons should match reconstruction quality.
-
The frame-prediction protocol is underspecified. Given the bidirectional video encoder, the construction of conditioning tokens using only observed frames needs clarification. Essential AR training and FVD evaluation settings are also missing.
[1] CAT: Content-Adaptive Image Tokenization
[2] One-D-Piece: Image Tokenizer Meets Quality-Controllable Compression
[3] VaporTok: RL-Driven Adaptive Video Tokenizer with Prior & Task Awareness
[4] Adaptive 1D Video Diffusion Autoencoder
Questions for authors
-
What specifically distinguishes this method from existing adaptive tokenizers, and which differences account for its improvements?
-
Do the gains persist against fixed-length and randomly allocated baselines at the same average token budget and training budget?
-
How is the rFVD stopping criterion implemented, and does the speedup persist against alternative search strategies at comparable quality?
-
How are K600 conditioning tokens constructed without access to future frames, and what are the evaluation settings?
I would consider raising my score if the authors adequately address the concerns above.