proomt

Search

Search posts, papers, and topics

All posts

Hugging Face Daily PapersSergey Kurinov, Alexey Upatov1 min readpaperintermediate

TAPe+ML: A Compact Structured Representation for Multi-Task Computer Vision

Read on arxiv.org

Summary

TAPe+ML v3 is a compact multi‑task vision system that replaces raw‑pixel processing with a structured TAPe representation. Using <100 k parameters, it achieves 84.7 mAP50 (65.3 mAP50‑95) on COCO detection, 80.7 mask mAP50 (58.4 mask mAP50‑95) on COCO segmentation, 92 % top‑1 on Imagenette and 89.9 % on ImageNet‑Real, while also showing robustness to distribution shift in video scene detection.

  • TAPe encodes relations among perceptual elements before recognition, allowing a shared representation across classification, detection, and segmentation.
  • The architecture is modular: background/contour processing → local object localization → prototype‑based classifier → coordinator for task‑specific sub‑models.
  • Parameter count stays under 100 k, dramatically reducing memory and compute compared to typical deep CNNs.
  • Empirical results on COCO and ImageNet show competitive accuracy despite the tiny model size.

If the TAPe representation can consistently shift modeling complexity from learned weights to structured inputs, it opens a path to ultra‑lightweight, multi‑task vision models suitable for edge devices and low‑bandwidth scenarios.

6/10
computer visionmulti taskcompact modelsstructured representation

Related reading

  1. Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies

    Distill frozen world‑model features into a Vision‑Language‑Action policy via a single alignment loss; no teacher at train time, no extra runtime cost. 0.8 B student runs 32 ms / 1.86 GB on RTX 5090, hits 97.9 % on LIBERO and improves RoboCasa‑GR1 from 48.2 % to 50.5 %, with transfer to real single‑arm and bimanual robots.

    Hugging Face Daily Papersarxiv.org1 minpaper

  2. Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash

    Cactus Needle 3 is a laddered attention LLM (29‑121 M params, CQ2‑bit quantised) that fits in 9‑29 MB binaries and runs on devices like a Raspberry Pi 5 (400‑4k t/s decode, 1‑10k t/s pre‑fill). It can perform tool‑calling, structured extraction, and text embedding locally, matching DeepSeek V4 Flash on downstream tasks after a single fine‑tune epoch. The post includes architecture details, benchm…

    Hacker News front pagecactuscompute.com7 minreleaseHN537185

  4. HuRo: Robotizing Human Videos for Scalable VLA Pretraining

    The paper introduces a pipeline that converts heterogeneous human videos into robot‑aligned observations and actions, creating the 630K‑episode HuRo dataset. Pretraining vision‑language‑action (VLA) policies on this data boosts real‑world manipulation success from ~51% to ~80% and improves out‑of‑distribution robustness.

    Hugging Face Daily Papersarxiv.org1 minpaper

  6. VC-Attention: Value Smoothing and Softmax Casting for Low-bit Attention

    VC-Attention introduces a training‑free low‑bit attention pipeline for diffusion transformers. It smooths value tensors via lightweight online clustering (V‑Smooth) and quantizes only the residual after subtracting block means, restoring the mean from the softmax row sum. It also replaces the FP32 softmax exponential with a fused FP8 cast (ExpCast‑FP8) that maps log‑scores directly to E4M3 probab…

    Hugging Face Daily Papersarxiv.org1 minpaper