Update README.md
Browse files
README.md
CHANGED
|
@@ -14,11 +14,11 @@ This work lays out that precise spatial understanding can emerge from simple poi
|
|
| 14 |
|
| 15 |
# Method
|
| 16 |
|
|
|
|
|
|
|
| 17 |
In this work, we explore the limits of MLLM pixel-level perception by predicting the next point in a contour with the simplest approach possible.
|
| 18 |
Without introducing any complex architectures or special patterns, we show how even minimalistic point prediction can achieve effective segmentation at the pixel level.
|
| 19 |
|
| 20 |
-
|
| 21 |
-
|
| 22 |
# Key Benefits
|
| 23 |
|
| 24 |
- **Simplicity**: SimpleSeg requires no specialized modules and adheres to the standard MLLM architecture, it can be seamlessly and efficiently integrated as a new, core pre-training task for foundation models, similar to visual grounding.
|
|
@@ -48,7 +48,6 @@ Without introducing any complex architectures or special patterns, we show how e
|
|
| 48 |
|
| 49 |
- **Referring Expression Comprehension** results
|
| 50 |
|
| 51 |
-
|
| 52 |
| Methods | refCOCO | | | refCOCO+ | | | refCOCOg | | Avg. |
|
| 53 |
|------------------|---------|----------|----------|----------|----------|----------|----------|----------|-------|
|
| 54 |
| | val | testA | testB | val | testA | testB | val | test | |
|
|
|
|
| 14 |
|
| 15 |
# Method
|
| 16 |
|
| 17 |
+
|
| 18 |
+
|
| 19 |
In this work, we explore the limits of MLLM pixel-level perception by predicting the next point in a contour with the simplest approach possible.
|
| 20 |
Without introducing any complex architectures or special patterns, we show how even minimalistic point prediction can achieve effective segmentation at the pixel level.
|
| 21 |
|
|
|
|
|
|
|
| 22 |
# Key Benefits
|
| 23 |
|
| 24 |
- **Simplicity**: SimpleSeg requires no specialized modules and adheres to the standard MLLM architecture, it can be seamlessly and efficiently integrated as a new, core pre-training task for foundation models, similar to visual grounding.
|
|
|
|
| 48 |
|
| 49 |
- **Referring Expression Comprehension** results
|
| 50 |
|
|
|
|
| 51 |
| Methods | refCOCO | | | refCOCO+ | | | refCOCOg | | Avg. |
|
| 52 |
|------------------|---------|----------|----------|----------|----------|----------|----------|----------|-------|
|
| 53 |
| | val | testA | testB | val | testA | testB | val | test | |
|