undefined | Better HN

0 pointsmagicalhippo2d ago0 comments

Just as example, here's what Qwen3.6 27B Q5_K_XL can do given this[1] image. I didn't do any prompt engineering here just a dead simple prompt: "Transcribe the following receipt. Put line items in a separate section, each line item separated by a double newline". Temperature set to 0.5.

Here's the output:

  Publix.
  Bradenton Commons Shopping Center
  4651 Cortez Rd. W.
  Bradenton, FL 34210
  Store Manager: Joe Galati
  941-792-7195
  
  N/O LF WHEAT BREAD 3.99 F
  
  PBX THCK L/S BACON 7.82 F
  
  PUBLIX BROWN GRAVY 0.83 F
  
  TOP SIRLOIN STEAK 11.74 F
  You Saved 3.92
  
  VITA PRTY SNK WINE 6.99 F
  You Saved 3.00
  
  ORGANIC CARROTS 1.69 F
  
  BRC FLRT EAT SMART 3.34 F
  1 @ 3 FOR 10.00
  You Saved 0.15
  
  GINGER ROOT 0.65 F
  0.13 lb @ 4.99/ lb
  
  POTATOES RUSSET 0.84 F
  0.65 lb @ 1.29/ lb
  
  POTATOES SWEET 0.49 F
  0.49 lb @ 0.99/ lb
  
  DELECT BSQUE CK/TN 10.99 T
  
  FS OUTSTRETCH UNSC 15.99 T
  
  Order Total 65.36
  Sales Tax 1.89
  Grand Total 67.25
  Credit Payment 67.25
  Change 0.00
  
  Savings Summary
  Special Price Savings 7.07
  ************************************************************
  * Your Savings at Publix *
  * 7.07 *
  ************************************************************
  
  Receipt ID: 5957 6249 2191 1277 712
  - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
  PRESTO!
  Trace #: 766630
  Reference #: 0098440513
  Acct #: XXXXXXXXXXXX2034
  Purchase VISA

[1]: https://i.pinimg.com/originals/41/08/dc/4108dcf51f15af464bb6...

0 comments

sebazzz2d ago

What is the difference between this and using normal OCR and then running that output through a LLM? It seems such a bazooka way to kill a fly to me using a modelime Qwen.

Doxin1d ago

For this example, perhaps not much, but a similar query also works. given this image[0] I queried qwen3.5-35b-a3b "where is waldo", and after thinking some got this back:

---

Waldo is located in the *center* of the image.

Here is how to spot him:

1. Look towards the middle of the beach scene.

2. Find the *red and white striped lifeguard stand/booth* on the right side of the center area.

3. Just to the left of that booth, there is a man standing up wearing his signature *red and white striped shirt*, blue jeans, and red hat with a pom-pom. He also has a camera around his neck.

---

Which is not generally something OCR can do. If you instead ask "how many horses are in this image?" you get this response:

---

Based on a careful look at the image, there are *3* equine animals (which appear to be donkeys or ponies used for beach rides).

1. One is near the top left, being ridden by someone in yellow shorts.

2. Right next to it is another one, being ridden by someone in blue and yellow.

3. On the far right edge of the image (near the water), there is a third grey animal being ridden by someone wearing a hat.

Now is this all anything you can't do with more boring machine learning? sure, but there's something incredibly convenient about how generic LLMs are. You don't need to train anything, just point the LLM at an image and ask.

[0] https://i.pinimg.com/originals/18/64/44/1864444c819a7adae742...

magicalhippoOP2d ago

For most tasks I agree. However once you've done your OCR you already have lost a lot of positional and context information, so for some tasks it might not be good enough.

If you have scanned PDFs that follow a template, like an invoice from a repeat supplier, then yeah OCR is definitely the way to go.

j / k navigate · click thread line to collapse