Pith. sign in

REVIEW 3 major objections 5 minor 76 references

Few-Shot Vision-Language Action-Incremental Policy Learning

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that a task-relation graph built from prompt similarities lets Transformer-based robot policies learn new manipulation tasks from one or five demonstrations while retaining old skills.

desk verdict Useful few-shot continual-learning recipe with big reported gains, but the CES update as written up-weights dissimilar tasks; until that sign/direction is fixed and code ships, the 25%+ gains can't be taken at face value. read the letter →

arxiv 2504.15517 v1 pith:5HHR3XYV submitted 2025-04-22 cs.RO cs.LG

Mingchen Song , Xiang Deng , Guoqiang Zhong , Qi Lv , Jia Wan , Yinchuan Li , Jianye Hao , Weili Guan This is my paper · ORCID
classification cs.ROcs.LG
keywords few-shotlearningincrementalcontinualroboticmanipulationimitationprompttaskrelationgraphcatastrophicforgetting
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces Few-Shot Action-Incremental Learning (FSAIL), a setting in which a robot policy is first trained on many demonstrations of base tasks and then must learn each new task from only one or five demonstrations while continuing to solve every earlier task. To solve it, the paper proposes TOPIC, a plug-in strategy for Transformer-based manipulation policies. TOPIC trains a small set of Task-Specific Prompts that are concatenated with language and multi-view visual tokens and processed by the policy's transformer, distilling a few demonstrations into task-specific features; it then builds a task relation graph from cosine distances between these prompts and uses the graph to mix previous tasks' policy weights into the new task's weights. Across the three base policies tested, TOPIC reports average accuracy improvements of 25.3, 28.2, and 26.4 percentage points in the 1-shot setting and 24.4, 29.4, and 25.0 points in the 5-shot setting, plus an 8.8-point average improvement in a real-world experiment. If the claims hold, data-hungry manipulation policies can be extended to new skills from a handful of demonstrations without large replay buffers or full retraining.

What carries the argument

The central object is a task relation graph whose nodes are task-specific prompts. For a new task $j$, the cosine distance $s_{ij}=d(\hat{P}_i,\hat{P}_j)$ between its prompt and every earlier prompt gives reuse coefficients, and the policy weights are updated as $\hat{W}_j=\lambda_1\left(\frac{1}{j-1}\sum_{i=1}^{j-1}s_{ij}W_i+W_j\right)+\lambda_2 W_{\mathrm{base}}$, with $\lambda_1$ and $\lambda_2$ balancing task-specific skills against the common base skill. The graph is what lets the method grow with the number of tasks: the same graph is extended at every session, and each newly learned skill becomes a reusable node. The Task-Specific Prompts do the representation work, turning a few demonstrations into a stable node identity despite the small sample size.

What would settle it

Randomize the reuse coefficients $s_{ij}$ in the weight-update equation while keeping the training procedure otherwise unchanged; if the average accuracy over sessions stays the same, the prompt-similarity graph is not what produces the reported gains.

Watch

Extended reading notes

Core claim

The central claim is that catastrophic forgetting in few-shot incremental manipulation can be avoided by making task identity explicit and by transferring old skills through a graph of task similarities. Each task gets a dedicated set of learnable Task-Specific Prompts that attend jointly to visual and language tokens inside the Multi-View Transformer encoder; after training, these prompts serve as compact descriptors of the task and are projected back onto the feature map to guide action prediction. When a new task arrives, its prompt is compared with all previous prompts by cosine distance, and the new policy weights are set to a weighted combination of the previous tasks' weights and the base-session weights. The paper reports that this two-part recipe, applied on top of three existing Transformer-based policies, roughly doubles their average accuracy across six sessions in both 1-shot and 5-shot FSAIL tasks, and it also improves average accuracy on a real mobile manipulation robot.

Load-bearing premise

The load-bearing premise is that a few demonstrations yield prompt vectors similar enough that cosine distance between them reflects reusable skills, so blending policy weights with those coefficients transfers skill rather than noise.

Editorial extensions

If this is right

  • Any Transformer manipulation policy that can expose its feature tokens can be fitted with TOPIC, since the text and visual encoders stay frozen and only prompts and policy weights are updated.
  • A robot does not need a replay buffer to remember old skills: the base weights and previous task weights are carried forward through the task relation graph rather than stored examples.
  • The more tasks accumulated in the graph, the more reusable skill nodes exist, so the method's capacity for transfer should grow with experience rather than saturating.
  • The reported 1-shot and 5-shot gains on three different policies indicate the benefit is not tied to one specific architecture, and the real-world result indicates the mechanism transfers beyond simulation, with a noticeable performance gap remaining.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The authors do not test whether the improvements survive randomizing the graph coefficients; if they do, CES would be acting mostly as a regularizer toward the base policy rather than as a true skill-transfer mechanism.
  • Because the graph uses cosine distance on prompts, a natural extension is to check whether the learned similarities align with compositional action and object structure, such as shared verbs like 'open' or shared objects like 'drawer', on a larger task inventory.
  • Since encoders are frozen, TOPIC should be composable with future, stronger backbone policies beyond the three tested, as long as its prompt projection and weight-mixing modules are retrained on their features.
  • The simulation-to-real gap reported in the paper suggests that the transfer gains depend partly on the visual encoder's robustness; a testable prediction is that stronger visual features shrink the gap.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper introduces Few-Shot Action-Incremental Learning (FSAIL), in which a vision-language Transformer policy is trained on a base session and then incrementally adapted to new manipulation tasks from one or five demonstrations, while being evaluated on all tasks seen so far. To address this, the authors propose TOPIC, which combines Task-Specific Prompts (TSP) that are concatenated with visual and language tokens in a multi-view Transformer, and a Continuous Evolution Strategy (CES) that constructs a task relation graph from the learned prompts and uses it to mix the policy weights of previous tasks. The method is integrated with RVT, SAM-E, and RVT-2 and evaluated on RLBench and a Mobile ALOHA robot against the base policies and replay-, regularization-, and S-prompts-based continual learning baselines. The paper reports large average success-rate improvements (up to about 28 points), together with ablations, coefficient sensitivity studies, a parameter/computation comparison, and a real-world experiment.

Significance. If the results hold, the contribution is significant: FSAIL is a practically relevant combination of few-shot and continual learning for robotic manipulation, and the modular integration with several strong Transformer policies makes the approach easy to adopt. The empirical apparatus is substantial, with five-run means and standard deviations, component ablations, hyperparameter sensitivity tables, per-task breakdowns, and a real-robot validation. I agree with the reader that the evaluation is not circular: the headline numbers are success rates against external baselines, not predictions from fitted parameters. However, the central CES update in Section III-D contains an internal sign inconsistency that must be resolved before the mechanism claims can be taken at face value.

major comments (3)
  1. [Section III-D, Eqs. (8)-(10)] There is a sign/direction inconsistency in the central skill-reuse mechanism. Eq. (8) defines s_ij as the cosine distance between task-specific prompts, which is small for similar tasks and large for dissimilar tasks. Eq. (9) then uses s_ij directly as the interpolation coefficient for previous weights W_i, so it up-weights dissimilar tasks and down-weights similar tasks. This contradicts the surrounding text, which says that the graph captures 'intrinsic relationships' and that skills are reused through the graph; it also contradicts Fig. 7 and Section IV-D1, where the authors compute cosine similarities and argue that similar tasks (e.g., 'open drawer' and 'put drawer') should be reused. Since the TSP & CES ablation in Table III shows that CES contributes a large part of the reported gain, this is not a cosmetic typo. The manuscript must either change Eq. (8)-(9) to use cosine similarity (or a monotone transform of the distance) or explicitly explain why dissimilar tasks are being up-weighted. The code repository is not yet available, so the actual implementation cannot be inspected to resolve the ambiguity.
  2. [Section IV-C, Tables IV-V and Fig. 5] The hyperparameters λ1, λ2, the number of task-specific prompts, and the prompt projection method are selected by ablations on the same benchmark and the same sessions that produce the headline results in Tables I and II. If these choices were made on the test sessions, the reported 25-29% improvements are partially selected on the test set. Please state explicitly whether a held-out validation set (or separate tasks) was used for these choices; if not, the claims should be tempered and the sensitivity of the headline results to these choices should be discussed.
  3. [Section III-C and Section IV-B] The paper does not specify how the correct task-specific prompt is selected at evaluation time. Each task has its own dedicated prompt, but the evaluation in session t includes test data from all tasks encountered so far; without a stated rule (e.g., task identity at test time, language-based prompt retrieval, or matching by the language instruction), the evaluation protocol is underspecified and the method cannot be reproduced from the description. Please clarify the prompt-selection mechanism and any assumptions about task identity.
minor comments (5)
  1. [Section IV-C3] The text says 'coefficients λ1 and λ2 in Equation 3', but the relevant update rule is Eq. (10); please correct the cross-reference.
  2. [Fig. 7 and Section IV-D1] The text and figure caption use 'cosine similarities', while Eq. (8) defines 'cosine distance'; please align the terminology throughout and state explicitly which quantity is used in Eq. (9).
  3. [Tables VII and VIII] The captions say 'Mean of 5 evaluations are reported' but no standard deviations are given, unlike Tables I and II; please either add standard deviations or explain why they are omitted in the per-task tables.
  4. [Sections IV-B, IV-C5, and V] The acronym is written as 'TPOIC' in several places (e.g., 'TPOIC remains notably superior', 'still performs better than the baseline model', and the conclusion); this typo should be fixed to 'TOPIC'.
  5. [Table VI and Section IV-C6] The table reports 'Params' as 35.6M for SAM-E and 35.5M for TOPIC, while the text emphasizes that TOPIC has fewer trainable parameters; please clarify whether the table reports total parameters or trainable parameters and how the frozen encoders are accounted for.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all central results are external benchmark measurements, and the flagged cosine-distance issue is a correctness concern rather than a derivation-by-construction.

full rationale

The paper's central claims are empirical success rates measured on RLBench and in real-world FSAIL tasks, and the reported 25-29% improvements are compared against external baselines (RVT, SAM-E, RVT-2) and standard continual-learning methods (replay, regularization, S-Prompts). No equation or table reduces those measured accuracy numbers to a quantity defined by TOPIC's own fitted parameters, so the main result is not circular. TSP and CES are architectural and heuristic components whose contributions are separately ablated in Table III; the lambda sweep in Table IV is hyperparameter selection on the same benchmark, which is an evaluation-protocol concern, not a circular derivation. The skeptical observation that Eq. 8 defines s_ij as cosine distance while Eq. 9 uses it as an interpolation coefficient (so dissimilar tasks receive larger weights) is an internal sign/direction inconsistency that could affect the correctness of the skill-reuse mechanism, but it does not make any predicted quantity equivalent by construction to an input, so it is not circularity. Section VI honestly lists limitations including computational resource constraints, a simulation-to-real gap, and incomplete resolution of catastrophic forgetting; these are scope statements, not admissions of circularity. The unreleased code limits reproducibility but does not create a circularity. Overall, the derivation chain is self-contained against external benchmarks, and no load-bearing step reduces to its own inputs.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The method introduces trainable prompt vectors and a task-relation graph as algorithmic components, not new physical entities. The main free parameters are the two balance coefficients, the prompt count, and the projection method, all selected by ablations on the same benchmark. The core load-bearing assumptions are that prompt similarity measures skill reusability and that base-session weights remain generally useful.

free parameters (4)
  • lambda1 = 0.2
    Chosen by ablation on the FSAIL benchmark (Table IV) to balance the contribution of task-specific weights in Equation 10.
  • lambda2 = 0.8
    Chosen by the same ablation to weight the base-session general skill Wbase.
  • number_of_task_specific_prompts = 5
    Selected by sweeping prompt counts on FSAIL tasks (Figure 5); no held-out validation is reported.
  • prompt_projection_method = Average Pooling
    Selected as best among identity, linear, MLP, and average pooling on the FSAIL benchmark (Table V).
assumptions (5)
  • domain assumption Skills are systematically reused across different tasks, e.g., pick-and-place appears in many tasks.
    Section III-D opens with this intuition; the CES weight interpolation in Equation 10 depends on it.
  • domain assumption Cosine distance between learned task-specific prompts reflects intrinsic task relationships.
    Equation 8 defines all relation coefficients by cosine distance; the paper provides only a qualitative similarity visualization as evidence.
  • domain assumption The base session yields a general skill weight Wbase that remains valid for all future tasks.
    Equation 10 includes lambda2 times Wbase in every new policy; if base skills are not general, new task policies inherit bias.
  • domain assumption Frozen visual and text encoders contain enough representation for unseen incremental tasks.
    The training procedure in Section III-E freezes both encoders during incremental sessions, assuming few-shot prompt tuning can adapt without encoder updates.
  • domain assumption Ten RLBench tasks provide a sufficient base for transfer to the five incremental tasks.
    The FSAIL benchmark is constructed on this assumption; stronger claims about general continual learning would require more diverse environments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Few-Shot Vision-Language Action-Incremental Policy Learning." pith.science (2026). https://pith.science/paper/5HHR3XYV

@misc{pith2026250415517,
  author       = {Pith},
  title        = {Pith review of: Few-Shot Vision-Language Action-Incremental Policy Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5HHR3XYV}},
  note         = {Machine review of arXiv:2504.15517}
}
abstract

Recently, Transformer-based robotic manipulation methods utilize multi-view spatial representations and language instructions to learn robot motion trajectories by leveraging numerous robot demonstrations. However, the collection of robot data is extremely challenging, and existing methods lack the capability for continuous learning on new tasks with only a few demonstrations. In this paper, we formulate these challenges as the Few-Shot Action-Incremental Learning (FSAIL) task, and accordingly design a Task-prOmpt graPh evolutIon poliCy (TOPIC) to address these issues. Specifically, to address the data scarcity issue in robotic imitation learning, TOPIC learns Task-Specific Prompts (TSP) through the deep interaction of multi-modal information within few-shot demonstrations, thereby effectively extracting the task-specific discriminative information. On the other hand, to enhance the capability for continual learning on new tasks and mitigate the issue of catastrophic forgetting, TOPIC adopts a Continuous Evolution Strategy (CES). CES leverages the intrinsic relationships between tasks to construct a task relation graph, which effectively facilitates the adaptation of new tasks by reusing skills learned from previous tasks. TOPIC pioneers few-shot continual learning in the robotic manipulation task, and extensive experimental results demonstrate that TOPIC outperforms state-of-the-art baselines by over 26$\%$ in success rate, significantly enhancing the continual learning capabilities of existing Transformer-based policies.

Figures

Figures reproduced from arXiv: 2504.15517 by the authors.

Figure 1
Figure 1. Illustration of our proposed Task-prOmpt graPh evo [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Comparisons of different Transformer-based policies and our proposed TOPIC. (a): Transformer-based policies include a [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The structure of our proposed Task-Specific Prompts [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: FSAIL Tasks in RLBench. We design 10 tasks in the base session and 5 tasks in the incremental session to validate the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Exploring the impact of the number of task-specific [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Exploring the impact of the number of tasks in the base [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 8
Figure 8. Figure 8: Real-World hardware configuration. We use the Cobot [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Qualitative Results in Real-world. We design 10 tasks in the base session and 5 tasks in the incremental session to [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

76 extracted references · 36 canonical work pages

  1. [1]

    V oxposer: Composable 3d value maps for robotic manipulation with language models,

    W. Huang, C. Wang, R. Zhang, Y . Li, J. Wu, and L. Fei-Fei, “V oxposer: Composable 3d value maps for robotic manipulation with language models,” arXiv preprint arXiv:2307.05973 , 2023

    arXiv 2023

  2. [2]

    Instruct2act: Mapping multi-modality instructions to robotic actions with large language model,

    S. Huang, Z. Jiang, H. Dong, Y . Qiao, P. Gao, and H. Li, “Instruct2act: Mapping multi-modality instructions to robotic actions with large language model,” arXiv preprint arXiv:2305.11176 , 2023

    arXiv 2023

  3. [3]

    Manipllm: Embodied multimodal large language model for object-centric robotic manipulation,

    X. Li, M. Zhang, Y . Geng, H. Geng, Y . Long, Y . Shen, R. Zhang, J. Liu, and H. Dong, “Manipllm: Embodied multimodal large language model for object-centric robotic manipulation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 18 061–18 070

    2024

  4. [4]

    Hierarchical diffusion policy for kinematics-aware multi-task robotic manipulation,

    X. Ma, S. Patidar, I. Haughton, and S. James, “Hierarchical diffusion policy for kinematics-aware multi-task robotic manipulation,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 18 081–18 090

    work page 2024

  5. [5]

    Dexterous grasp transformer,

    G.-H. Xu, Y .-L. Wei, D. Zheng, X.-M. Wu, and W.-S. Zheng, “Dexterous grasp transformer,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 17 933–17 942

    work page 2024

  6. [6]

    Unidexgrasp: Universal robotic dexterous grasping via learning diverse proposal generation and goal-conditioned policy,

    Y . Xu, W. Wan, J. Zhang, H. Liu, Z. Shan, H. Shen, R. Wang, H. Geng, Y . Weng, J. Chen et al. , “Unidexgrasp: Universal robotic dexterous grasping via learning diverse proposal generation and goal-conditioned policy,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 4737–4746

    work page 2023

  7. [7]

    Palm-e: An embodied multimodal language model,

    D. Driess, F. Xia, M. S. Sajjadi, C. Lynch, A. Chowdhery, B. Ichter, A. Wahid, J. Tompson, Q. Vuong, T. Yu et al., “Palm-e: An embodied multimodal language model,” arXiv preprint arXiv:2303.03378 , 2023

    arXiv 2023

  8. [8]

    Vima: General robot manipulation with multimodal prompts,

    Y . Jiang, A. Gupta, Z. Zhang, G. Wang, Y . Dou, Y . Chen, L. Fei-Fei, A. Anandkumar, Y . Zhu, and L. Fan, “Vima: General robot manipulation with multimodal prompts,” arXiv preprint arXiv:2210.03094, vol. 2, no. 3, p. 6, 2022

    arXiv

Show all 76 references
  1. [9]

    Mastering robot manipulation with multimodal prompts through pretraining and multi-task fine-tuning,

    J. Li, Q. Gao, M. Johnston, X. Gao, X. He, S. Shakiah, H. Shi, R. Ghanadan, and W. Y . Wang, “Mastering robot manipulation with multimodal prompts through pretraining and multi-task fine-tuning,”arXiv preprint arXiv:2310.09676, 2023

    arXiv

  2. [10]

    Instruction-driven history-aware policies for robotic manipulations,

    P.-L. Guhur, S. Chen, R. G. Pinel, M. Tapaswi, I. Laptev, and C. Schmid, “Instruction-driven history-aware policies for robotic manipulations,” in Conference on Robot Learning . PMLR, 2023, pp. 175–187

    2023

  3. [11]

    Rvt: Robotic view transformer for 3d object manipulation,

    A. Goyal, J. Xu, Y . Guo, V . Blukis, Y .-W. Chao, and D. Fox, “Rvt: Robotic view transformer for 3d object manipulation,” in Conference on Robot Learning. PMLR, 2023, pp. 694–710

    2023

  4. [12]

    Rvt-2: Learning precise manipulation from few demonstrations,

    A. Goyal, V . Blukis, J. Xu, Y . Guo, Y .-W. Chao, and D. Fox, “Rvt-2: Learning precise manipulation from few demonstrations,” arXiv preprint arXiv:2406.08545, 2024

    2024 arXiv

  5. [13]

    Sam-e: Leveraging visual foundation model with sequence imitation for embodied manipulation,

    J. Zhang, C. Bai, H. He, W. Xia, Z. Wang, B. Zhao, X. Li, and X. Li, “Sam-e: Leveraging visual foundation model with sequence imitation for embodied manipulation,” arXiv preprint arXiv:2405.19586 , 2024

    2024 arXiv

  6. [14]

    Sugar: Pre-training 3d visual representations for robotics,

    S. Chen, R. Garcia, I. Laptev, and C. Schmid, “Sugar: Pre-training 3d visual representations for robotics,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 18 049–18 060

    2024

  7. [15]

    Semantic prompt for few-shot image recognition,

    W. Chen, C. Si, Z. Zhang, L. Wang, Z. Wang, and T. Tan, “Semantic prompt for few-shot image recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 23 581–23 591

    2023

  8. [16]

    Winclip: Zero-/few-shot anomaly classification and segmentation,

    J. Jeong, Y . Zou, T. Kim, D. Zhang, A. Ravichandran, and O. Dabeer, “Winclip: Zero-/few-shot anomaly classification and segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 19 606–19 616

    2023

  9. [17]

    Not all features matter: Enhancing few-shot clip with adaptive prior refinement,

    X. Zhu, R. Zhang, B. He, A. Zhou, D. Wang, B. Zhao, and P. Gao, “Not all features matter: Enhancing few-shot clip with adaptive prior refinement,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 2605–2615

    2023

  10. [18]

    Tip-adapter: Training-free adaption of clip for few-shot classification,

    R. Zhang, W. Zhang, R. Fang, P. Gao, K. Li, J. Dai, Y . Qiao, and H. Li, “Tip-adapter: Training-free adaption of clip for few-shot classification,” in European conference on computer vision . Springer, 2022, pp. 493–510

    2022

  11. [19]

    Conditional prompt learning for vision-language models,

    K. Zhou, J. Yang, C. C. Loy, and Z. Liu, “Conditional prompt learning for vision-language models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2022, pp. 16 816–16 825

    2022

  12. [20]

    Smae: Few-shot learning for hdr deghosting with saturation-aware masked autoencoders,

    Q. Yan, S. Zhang, W. Chen, H. Tang, Y . Zhu, J. Sun, L. Van Gool, and Y . Zhang, “Smae: Few-shot learning for hdr deghosting with saturation-aware masked autoencoders,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 5775–5784

    2023

  13. [21]

    Styleadv: Meta style adversarial training for cross-domain few-shot learning,

    Y . Fu, Y . Xie, Y . Fu, and Y .-G. Jiang, “Styleadv: Meta style adversarial training for cross-domain few-shot learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 24 575–24 584

    2023

  14. [22]

    Rethinking few-shot medical segmentation: a vector quantization view,

    S. Huang, T. Xu, N. Shen, F. Mu, and J. Li, “Rethinking few-shot medical segmentation: a vector quantization view,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 3072–3081

    2023

  15. [23]

    Hyperbolic insights with knowledge distillation for cross-domain few-shot learning,

    X. Yang, D. Kong, N. Wang, and X. Gao, “Hyperbolic insights with knowledge distillation for cross-domain few-shot learning,” IEEE Transactions on Image Processing , 2025

    2025

  16. [24]

    Cross-modal contrastive learning network for few-shot action recognition,

    X. Wang, Y . Yan, H.-M. Hu, B. Li, and H. Wang, “Cross-modal contrastive learning network for few-shot action recognition,” IEEE Transactions on Image Processing , vol. 33, pp. 1257–1271, 2024

    2024

  17. [25]

    Transductive few-shot learning with enhanced spectral-spatial embedding for hyperspectral image classification,

    B. Xi, Y . Zhang, J. Li, Y . Huang, Y . Li, Z. Li, and J. Chanussot, “Transductive few-shot learning with enhanced spectral-spatial embedding for hyperspectral image classification,” IEEE Transactions on Image Processing, 2025

    2025

  18. [26]

    Enhancing information maximization with distance-aware contrastive learning for source-free cross-domain few-shot learning,

    H. Xu, L. Liu, S. Zhi, S. Fu, Z. Su, M.-M. Cheng, and Y . Liu, “Enhancing information maximization with distance-aware contrastive learning for source-free cross-domain few-shot learning,” IEEE Transactions on Image Processing, 2024

    2024

  19. [27]

    One-shot imitation learning,

    Y . Duan, M. Andrychowicz, B. Stadie, O. Jonathan Ho, J. Schneider, I. Sutskever, P. Abbeel, and W. Zaremba, “One-shot imitation learning,” Advances in neural information processing systems , vol. 30, 2017

    2017

  20. [28]

    One-shot visual imitation learning via meta-learning,

    C. Finn, T. Yu, T. Zhang, P. Abbeel, and S. Levine, “One-shot visual imitation learning via meta-learning,” in Conference on robot learning . PMLR, 2017, pp. 357–368

    2017

  21. [29]

    Task-embedded control networks for few-shot imitation learning,

    S. James, M. Bloesch, and A. J. Davison, “Task-embedded control networks for few-shot imitation learning,” in Conference on robot learning. PMLR, 2018, pp. 783–795

    2018

  22. [30]

    One-shot imitation learning with invariance matching for robotic manipulation,

    X. Zhang and A. Boularias, “One-shot imitation learning with invariance matching for robotic manipulation,” arXiv preprint arXiv:2405.13178 , 2024

    2024 arXiv

  23. [31]

    R3m: A universal visual representation for robot manipulation,

    S. Nair, A. Rajeswaran, V . Kumar, C. Finn, and A. Gupta, “R3m: A universal visual representation for robot manipulation,” arXiv preprint arXiv:2203.12601, 2022

    2022 arXiv

  24. [32]

    Robocat: A self-improving generalist agent for robotic manipulation,

    K. Bousmalis, G. Vezzani, D. Rao, C. M. Devin, A. X. Lee, M. B. Villalonga, T. Davchev, Y . Zhou, A. Gupta, A. Raju et al., “Robocat: A self-improving generalist agent for robotic manipulation,” Transactions on Machine Learning Research , 2023

    2023

  25. [33]

    Octo: An open-source generalist robot policy,

    O. M. Team, D. Ghosh, H. Walke, K. Pertsch, K. Black, O. Mees, S. Dasari, J. Hejna, T. Kreiman, C. Xu et al., “Octo: An open-source generalist robot policy,” arXiv preprint arXiv:2405.12213 , 2024

    2024 arXiv

  26. [34]

    Openvla: An open- source vision-language-action model,

    M. J. Kim, K. Pertsch, S. Karamcheti, T. Xiao, A. Balakrishna, S. Nair, R. Rafailov, E. Foster, G. Lam, P. Sanketi et al., “Openvla: An open- source vision-language-action model,” arXiv preprint arXiv:2406.09246 , 2024

    2024 arXiv

  27. [35]

    Perceiver-actor: A multi-task transformer for robotic manipulation,

    M. Shridhar, L. Manuelli, and D. Fox, “Perceiver-actor: A multi-task transformer for robotic manipulation,” in Conference on Robot Learning . PMLR, 2023, pp. 785–799

    2023

  28. [36]

    Gnfactor: Multi-task real robot learning with generalizable neural feature fields,

    Y . Ze, G. Yan, Y .-H. Wu, A. Macaluso, Y . Ge, J. Ye, N. Hansen, L. E. Li, and X. Wang, “Gnfactor: Multi-task real robot learning with generalizable neural feature fields,” in Conference on Robot Learning . PMLR, 2023, pp. 284–301

    2023

  29. [37]

    Revisiting class-incremental learning with pre-trained models: Generalizability and adaptivity are all you need,

    D.-W. Zhou, Z.-W. Cai, H.-J. Ye, D.-C. Zhan, and Z. Liu, “Revisiting class-incremental learning with pre-trained models: Generalizability and adaptivity are all you need,” arXiv preprint arXiv:2303.07338 , 2023

    2023 arXiv

  30. [38]

    Foster: Feature boosting and compression for class-incremental learning,

    F.-Y . Wang, D.-W. Zhou, H.-J. Ye, and D.-C. Zhan, “Foster: Feature boosting and compression for class-incremental learning,” in European conference on computer vision . Springer, 2022, pp. 398–414

    2022

  31. [39]

    Topology-preserving class-incremental learning,

    X. Tao, X. Chang, X. Hong, X. Wei, and Y . Gong, “Topology-preserving class-incremental learning,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIX 16. Springer, 2020, pp. 254–270

    2020

  32. [40]

    Mamba- fscil: Dynamic adaptation with selective state space model for few-shot class-incremental learning,

    X. Li, Y . Yang, J. Wu, B. Ghanem, L. Nie, and M. Zhang, “Mamba- fscil: Dynamic adaptation with selective state space model for few-shot class-incremental learning,” arXiv preprint arXiv:2407.06136 , 2024

    2024 arXiv

  33. [41]

    Multimodal parameter-efficient few-shot class incremental learning,

    M. D’Alessandro, A. Alonso, E. Calabr ´es, and M. Galar, “Multimodal parameter-efficient few-shot class incremental learning,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3393–3403

    2023

  34. [42]

    Few-shot class-incremental learning,

    X. Tao, X. Hong, X. Chang, S. Dong, X. Wei, and Y . Gong, “Few-shot class-incremental learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 12 183–12 192

    2020

  35. [43]

    Few-shot incremental learning with continually evolved classifiers,

    C. Zhang, N. Song, G. Lin, Y . Zheng, P. Pan, and Y . Xu, “Few-shot incremental learning with continually evolved classifiers,” in Proceedings 13 of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 12 455–12 464

    2021

  36. [44]

    Forward compatible few-shot class-incremental learning,

    D.-W. Zhou, F.-Y . Wang, H.-J. Ye, L. Ma, S. Pu, and D.-C. Zhan, “Forward compatible few-shot class-incremental learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2022, pp. 9046–9056

    2022

  37. [45]

    Gkeal: Gaussian kernel embedded analytic learning for few-shot class incremental task,

    H. Zhuang, Z. Weng, R. He, Z. Lin, and Z. Zeng, “Gkeal: Gaussian kernel embedded analytic learning for few-shot class incremental task,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 7746–7755

    2023

  38. [46]

    Der: Dynamically expandable representation for class incremental learning,

    S. Yan, J. Xie, and X. He, “Der: Dynamically expandable representation for class incremental learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 3014– 3023

    2021

  39. [47]

    Ntk-guided few-shot class incremental learning,

    J. Liu, Z. Ji, Y . Pang, and Y . Yu, “Ntk-guided few-shot class incremental learning,” IEEE Transactions on Image Processing , 2024

    2024

  40. [48]

    Relationship-incremental scene graph generation by a divide-and-conquer pipeline with feature adapter,

    X. Li, G. Zheng, Y . Yu, N. Ji, and X. Li, “Relationship-incremental scene graph generation by a divide-and-conquer pipeline with feature adapter,” IEEE Transactions on Image Processing , 2024

    2024

  41. [49]

    Memorizing complementation network for few-shot class-incremental learning,

    Z. Ji, Z. Hou, X. Liu, Y . Pang, and X. Li, “Memorizing complementation network for few-shot class-incremental learning,” IEEE Transactions on Image Processing, vol. 32, pp. 937–948, 2023

    2023

  42. [50]

    Overcoming catastrophic forgetting in neural networks,

    J. Kirkpatrick, R. Pascanu, N. Rabinowitz, J. Veness, G. Desjardins, A. A. Rusu, K. Milan, J. Quan, T. Ramalho, A. Grabska-Barwinska et al., “Overcoming catastrophic forgetting in neural networks,” Proceedings of the national academy of sciences , vol. 114, no. 13, pp. 3521–3526, 2017

    2017

  43. [51]

    Memory aware synapses: Learning what (not) to forget,

    R. Aljundi, F. Babiloni, M. Elhoseiny, M. Rohrbach, and T. Tuytelaars, “Memory aware synapses: Learning what (not) to forget,” in Proceedings of the European conference on computer vision (ECCV) , 2018, pp. 139– 154

    2018

  44. [52]

    Continual learning through synaptic intelligence,

    F. Zenke, B. Poole, and S. Ganguli, “Continual learning through synaptic intelligence,” in International conference on machine learning . PMLR, 2017, pp. 3987–3995

    2017

  45. [53]

    icarl: Incremental classifier and representation learning,

    S.-A. Rebuffi, A. Kolesnikov, G. Sperl, and C. H. Lampert, “icarl: Incremental classifier and representation learning,” in Proceedings of the IEEE conference on Computer Vision and Pattern Recognition , 2017, pp. 2001–2010

    2017

  46. [54]

    Memory-efficient incremental learning through feature adaptation,

    A. Iscen, J. Zhang, S. Lazebnik, and C. Schmid, “Memory-efficient incremental learning through feature adaptation,” in Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVI 16 . Springer, 2020, pp. 699–715

    2020

  47. [55]

    Incremental learning using conditional adversarial networks,

    Y . Xiang, Y . Fu, P. Ji, and H. Huang, “Incremental learning using conditional adversarial networks,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2019, pp. 6619–6628

    2019

  48. [56]

    Adaptive memory replay for continual learning,

    J. S. Smith, L. Valkov, S. Halbe, V . Gutta, R. Feris, Z. Kira, and L. Karlinsky, “Adaptive memory replay for continual learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 3605–3615

    2024

  49. [57]

    Balanced destruction- reconstruction dynamics for memory-replay class incremental learning,

    Y . Zhou, J. Yao, F. Hong, Y . Zhang, and Y . Wang, “Balanced destruction- reconstruction dynamics for memory-replay class incremental learning,” IEEE Transactions on Image Processing , 2024

    2024

  50. [58]

    Coda-prompt: Continual decomposed attention-based prompting for rehearsal-free continual learning,

    J. S. Smith, L. Karlinsky, V . Gutta, P. Cascante-Bonilla, D. Kim, A. Arbelle, R. Panda, R. Feris, and Z. Kira, “Coda-prompt: Continual decomposed attention-based prompting for rehearsal-free continual learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and...

    2023

  51. [59]

    Hierarchical decomposition of prompt-based continual learning: Rethinking obscured sub-optimality,

    L. Wang, J. Xie, X. Zhang, M. Huang, H. Su, and J. Zhu, “Hierarchical decomposition of prompt-based continual learning: Rethinking obscured sub-optimality,” Advances in Neural Information Processing Systems , vol. 36, 2024

    2024

  52. [60]

    Dualprompt: Complementary prompting for rehearsal-free continual learning,

    Z. Wang, Z. Zhang, S. Ebrahimi, R. Sun, H. Zhang, C.-Y . Lee, X. Ren, G. Su, V . Perot, J. Dyet al., “Dualprompt: Complementary prompting for rehearsal-free continual learning,” in European Conference on Computer Vision. Springer, 2022, pp. 631–648

    2022

  53. [61]

    S-prompts learning with pre- trained transformers: An occam’s razor for domain incremental learning,

    Y . Wang, Z. Huang, and X. Hong, “S-prompts learning with pre- trained transformers: An occam’s razor for domain incremental learning,” Advances in Neural Information Processing Systems , vol. 35, pp. 5682– 5695, 2022

    2022

  54. [62]

    Few-shot continual active learning by a robot,

    A. Ayub and C. Fendley, “Few-shot continual active learning by a robot,” Advances in Neural Information Processing Systems , vol. 35, pp. 30 612– 30 624, 2022

    2022

  55. [63]

    Vision-language navigation with continual learning,

    Z. Li, Y . Lv, Z. Tu, D. Shang, and H. Qiao, “Vision-language navigation with continual learning,” arXiv preprint arXiv:2409.02561 , 2024

    2024 arXiv

  56. [64]

    Continual vision-and-language navigation,

    S. Jeong, G.-C. Kang, S. Choi, J. Kim, and B.-T. Zhang, “Continual vision-and-language navigation,” arXiv preprint arXiv:2403.15049 , 2024

    2024

  57. [65]

    Libero: Benchmarking knowledge transfer for lifelong robot learning,

    B. Liu, Y . Zhu, C. Gao, Y . Feng, Q. Liu, Y . Zhu, and P. Stone, “Libero: Benchmarking knowledge transfer for lifelong robot learning,” Advances in Neural Information Processing Systems , vol. 36, pp. 44 776–44 791, 2023

    2023

  58. [66]

    Lotus: Continual imitation learning for robot manipulation through unsupervised skill discovery,

    W. Wan, Y . Zhu, R. Shah, and Y . Zhu, “Lotus: Continual imitation learning for robot manipulation through unsupervised skill discovery,” in 2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 537–544

    2024

  59. [67]

    M2distill: Multi-modal distillation for lifelong imitation learning,

    K. Roy, A. Dissanayake, B. Tidd, and P. Moghadam, “M2distill: Multi-modal distillation for lifelong imitation learning,” arXiv preprint arXiv:2410.00064, 2024

    2024 arXiv

  60. [68]

    Tail: Task-specific adapters for imitation learning with large pretrained models,

    Z. Liu, J. Zhang, K. Asadi, Y . Liu, D. Zhao, S. Sabach, and R. Fakoor, “Tail: Task-specific adapters for imitation learning with large pretrained models,” arXiv preprint arXiv:2310.05905 , 2023

    2023 arXiv

  61. [69]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,” Advances in Neural Information Processing Systems, 2017

    2017

  62. [70]

    Sparse diffusion policy: A sparse, reusable, and flexible policy for robot learning,

    Y . Wang, Y . Zhang, M. Huo, R. Tian, X. Zhang, Y . Xie, C. Xu, P. Ji, W. Zhan, M. Ding et al., “Sparse diffusion policy: A sparse, reusable, and flexible policy for robot learning,” arXiv preprint arXiv:2407.01531 , 2024

    2024 arXiv

  63. [71]

    Vision-language foundation models as effective robot imitators,

    X. Li, M. Liu, H. Zhang, C. Yu, J. Xu, H. Wu, C. Cheang, Y . Jing, W. Zhang, H. Liu et al., “Vision-language foundation models as effective robot imitators,” in International Conference on Learning Representations, 2024

    2024

  64. [72]

    Rlbench: The robot learning benchmark & learning environment,

    S. James, Z. Ma, D. R. Arrojo, and A. J. Davison, “Rlbench: The robot learning benchmark & learning environment,” IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 3019–3026, 2020

    2020

  65. [73]

    Lora: Low-rank adaptation of large language models,

    E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” arXiv preprint arXiv:2106.09685 , 2021

    2021 arXiv

  66. [74]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Lo et al., “Segment anything,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 4015–4026

    2023

  67. [75]

    Mobile aloha: Learning bimanual mobile manipulation with low-cost whole-body teleoperation,

    Z. Fu, T. Z. Zhao, and C. Finn, “Mobile aloha: Learning bimanual mobile manipulation with low-cost whole-body teleoperation,” arXiv preprint arXiv:2401.02117, 2024

    2024 arXiv

  68. [76]

    Learning fine- grained bimanual manipulation with low-cost hardware,

    T. Z. Zhao, V . Kumar, S. Levine, and C. Finn, “Learning fine- grained bimanual manipulation with low-cost hardware,” arXiv preprint arXiv:2304.13705, 2023

    2023 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.